Lesson-05-course.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Search Tree -> Similar Words"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "csv_path = './data/sqlResult_1558435.csv'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "os.path.exists(csv_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "content = pd.read_csv(csv_path, encoding='gb18030')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "content = content.fillna('')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>author</th>\n",
       "      <th>source</th>\n",
       "      <th>content</th>\n",
       "      <th>feature</th>\n",
       "      <th>title</th>\n",
       "      <th>url</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>89617</td>\n",
       "      <td></td>\n",
       "      <td>快科技@http://www.kkj.cn/</td>\n",
       "      <td>此外，自本周（6月12日）起，除小米手机6等15款机型外，其余机型已暂停更新发布（含开发版/...</td>\n",
       "      <td>{\"type\":\"科技\",\"site\":\"cnbeta\",\"commentNum\":\"37\"...</td>\n",
       "      <td>小米MIUI 9首批机型曝光：共计15款</td>\n",
       "      <td>http://www.cnbeta.com/articles/tech/623597.htm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>89616</td>\n",
       "      <td></td>\n",
       "      <td>快科技@http://www.kkj.cn/</td>\n",
       "      <td>骁龙835作为唯一通过Windows 10桌面平台认证的ARM处理器，高通强调，不会因为只考...</td>\n",
       "      <td>{\"type\":\"科技\",\"site\":\"cnbeta\",\"commentNum\":\"15\"...</td>\n",
       "      <td>骁龙835在Windows 10上的性能表现有望改善</td>\n",
       "      <td>http://www.cnbeta.com/articles/tech/623599.htm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>89615</td>\n",
       "      <td></td>\n",
       "      <td>快科技@http://www.kkj.cn/</td>\n",
       "      <td>此前的一加3T搭载的是3400mAh电池，DashCharge快充规格为5V/4A。\\r\\n...</td>\n",
       "      <td>{\"type\":\"科技\",\"site\":\"cnbeta\",\"commentNum\":\"18\"...</td>\n",
       "      <td>一加手机5细节曝光：3300mAh、充半小时用1天</td>\n",
       "      <td>http://www.cnbeta.com/articles/tech/623601.htm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>89614</td>\n",
       "      <td></td>\n",
       "      <td>新华社</td>\n",
       "      <td>这是6月18日在葡萄牙中部大佩德罗冈地区拍摄的被森林大火烧毁的汽车。新华社记者张立云摄\\r\\n</td>\n",
       "      <td>{\"type\":\"国际新闻\",\"site\":\"环球\",\"commentNum\":\"0\",\"j...</td>\n",
       "      <td>葡森林火灾造成至少62人死亡 政府宣布进入紧急状态（组图）</td>\n",
       "      <td>http://world.huanqiu.com/hot/2017-06/10866126....</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>89613</td>\n",
       "      <td>胡淑丽_MN7479</td>\n",
       "      <td>深圳大件事</td>\n",
       "      <td>（原标题：44岁女子跑深圳约会网友被拒，暴雨中裸身奔走……）\\r\\n@深圳交警微博称：昨日清...</td>\n",
       "      <td>{\"type\":\"新闻\",\"site\":\"网易热门\",\"commentNum\":\"978\",...</td>\n",
       "      <td>44岁女子约网友被拒暴雨中裸奔 交警为其披衣相随</td>\n",
       "      <td>http://news.163.com/17/0618/00/CN617P3Q0001875...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      id      author                  source  \\\n",
       "0  89617              快科技@http://www.kkj.cn/   \n",
       "1  89616              快科技@http://www.kkj.cn/   \n",
       "2  89615              快科技@http://www.kkj.cn/   \n",
       "3  89614                                 新华社   \n",
       "4  89613  胡淑丽_MN7479                   深圳大件事   \n",
       "\n",
       "                                             content  \\\n",
       "0  此外，自本周（6月12日）起，除小米手机6等15款机型外，其余机型已暂停更新发布（含开发版/...   \n",
       "1  骁龙835作为唯一通过Windows 10桌面平台认证的ARM处理器，高通强调，不会因为只考...   \n",
       "2  此前的一加3T搭载的是3400mAh电池，DashCharge快充规格为5V/4A。\\r\\n...   \n",
       "3    这是6月18日在葡萄牙中部大佩德罗冈地区拍摄的被森林大火烧毁的汽车。新华社记者张立云摄\\r\\n   \n",
       "4  （原标题：44岁女子跑深圳约会网友被拒，暴雨中裸身奔走……）\\r\\n@深圳交警微博称：昨日清...   \n",
       "\n",
       "                                             feature  \\\n",
       "0  {\"type\":\"科技\",\"site\":\"cnbeta\",\"commentNum\":\"37\"...   \n",
       "1  {\"type\":\"科技\",\"site\":\"cnbeta\",\"commentNum\":\"15\"...   \n",
       "2  {\"type\":\"科技\",\"site\":\"cnbeta\",\"commentNum\":\"18\"...   \n",
       "3  {\"type\":\"国际新闻\",\"site\":\"环球\",\"commentNum\":\"0\",\"j...   \n",
       "4  {\"type\":\"新闻\",\"site\":\"网易热门\",\"commentNum\":\"978\",...   \n",
       "\n",
       "                           title  \\\n",
       "0           小米MIUI 9首批机型曝光：共计15款   \n",
       "1     骁龙835在Windows 10上的性能表现有望改善   \n",
       "2      一加手机5细节曝光：3300mAh、充半小时用1天   \n",
       "3  葡森林火灾造成至少62人死亡 政府宣布进入紧急状态（组图）   \n",
       "4       44岁女子约网友被拒暴雨中裸奔 交警为其披衣相随   \n",
       "\n",
       "                                                 url  \n",
       "0     http://www.cnbeta.com/articles/tech/623597.htm  \n",
       "1     http://www.cnbeta.com/articles/tech/623599.htm  \n",
       "2     http://www.cnbeta.com/articles/tech/623601.htm  \n",
       "3  http://world.huanqiu.com/hot/2017-06/10866126....  \n",
       "4  http://news.163.com/17/0618/00/CN617P3Q0001875...  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "content.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "news_content = content['content'].tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "import jieba"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "def cut(string): return ' '.join(jieba.cut(string))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'这是 一个 测试'"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cut('这是一个测试')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "import re"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "def token(string):\n",
    "    return re.findall(r'[\\d|\\w]+', string)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['这是一个测试']"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "token('这是一个测试\\n\\n\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "news_content = [token(n) for n in news_content]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "89611"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(news_content)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['此外', '自本周', '6月12日', '起', '除小米手机6等15款机型外', '其余机型已暂停更新发布', '含开发版', '体验版内测', '稳定版暂不受影响', '以确保工程师可以集中全部精力进行系统优化工作', '有人猜测这也是将精力主要用到MIUI', '9的研发之中', 'MIUI', '8去年5月发布', '距今已有一年有余', '也是时候更新换代了', '当然', '关于MIUI', '9的确切信息', '我们还是等待官方消息']\n"
     ]
    }
   ],
   "source": [
    "print(news_content[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "news_content = [' '.join(n) for n in news_content]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['此外 自本周 6月12日 起 除小米手机6等15款机型外 其余机型已暂停更新发布 含开发版 体验版内测 稳定版暂不受影响 以确保工程师可以集中全部精力进行系统优化工作 有人猜测这也是将精力主要用到MIUI 9的研发之中 MIUI 8去年5月发布 距今已有一年有余 也是时候更新换代了 当然 关于MIUI 9的确切信息 我们还是等待官方消息']"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_content[:1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'此前 的 一加 3T 搭载 的 是 3400mAh 电池   DashCharge 快充 规格 为 5V   4A   至于 电池 缩水   可能 与 刘作 虎 所说   一加 手机 5 要 做 市面 最 轻薄 大屏 旗舰 的 设定 有关   按照 目前 掌握 的 资料   一加 手机 5 拥有 5   5 寸 1080P 三星 AMOLED 显示屏   6G   8GB   RAM   64GB   128GB   ROM   双 1600 万 摄像头   备货 量   惊喜   根据 京东 泄露 的 信息   一加 5 起 售价 是 xx99 元   应该 是 在 2799   2899   2999 中 的 某个'"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cut(news_content[2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "news_content = [cut(n) for n in news_content]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "89611"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(news_content)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'此外   自 本周   6 月 12 日   起   除 小米 手机 6 等 15 款 机型 外   其余 机型 已 暂停 更新 发布   含 开发 版   体验版 内测   稳定版 暂不受 影响   以 确保 工程师 可以 集中 全部 精力 进行 系统优化 工作   有人 猜测 这 也 是 将 精力 主要 用到 MIUI   9 的 研发 之中   MIUI   8 去年 5 月 发布   距今已有 一年 有余   也 是 时候 更新换代 了   当然   关于 MIUI   9 的 确切 信息   我们 还是 等待 官方消息'"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_content[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('news-sentences-cut.txt', 'w', encoding='utf8') as f:\n",
    "    for n in news_content:\n",
    "        f.write(n + '\\n')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "from gensim.models import Word2Vec"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "from gensim.models.word2vec import LineSentence"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "Word2Vec?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "news_word2ve= Word2Vec(LineSentence('news-sentences-cut.txt'), size=35, workers=8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\wanga26\\AppData\\Local\\Continuum\\anaconda3\\envs\\cs231n\\lib\\site-packages\\ipykernel\\__main__.py:1: DeprecationWarning: Call to deprecated `most_similar` (Method will be removed in 4.0.0, use self.wv.most_similar() instead).\n",
      "  if __name__ == '__main__':\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[('捷克', 0.8672915101051331),\n",
       " ('意大利', 0.8525253534317017),\n",
       " ('克罗地亚', 0.8454985022544861),\n",
       " ('巴塞罗那', 0.8235341906547546),\n",
       " ('瑞士', 0.8135308027267456),\n",
       " ('摩洛哥', 0.8098884224891663),\n",
       " ('乌拉圭', 0.8071278929710388),\n",
       " ('索', 0.8041039705276489),\n",
       " ('丹麦', 0.8013482093811035),\n",
       " ('西班牙', 0.8000324964523315),\n",
       " ('罗马尼亚', 0.7968207597732544),\n",
       " ('奥地利', 0.7957209944725037),\n",
       " ('马德里', 0.7858061790466309),\n",
       " ('拉脱维亚', 0.7844109535217285),\n",
       " ('比什凯克', 0.7844077348709106),\n",
       " ('比利时', 0.7838754653930664),\n",
       " ('科特迪瓦', 0.7754402160644531),\n",
       " ('巴拉圭', 0.7748613953590393),\n",
       " ('黑文市', 0.7727378010749817),\n",
       " ('斯洛文尼亚', 0.7718468904495239)]"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_word2ve.most_similar('葡萄牙', topn=20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## More Date, Better Results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. 分词的问题\n",
    "2. **数据量**，数据越多，效果越好，维基百科加进来，那么同义词就要好很多。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\wanga26\\AppData\\Local\\Continuum\\anaconda3\\envs\\cs231n\\lib\\site-packages\\ipykernel\\__main__.py:1: DeprecationWarning: Call to deprecated `most_similar` (Method will be removed in 4.0.0, use self.wv.most_similar() instead).\n",
      "  if __name__ == '__main__':\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[('罗马尼亚', 0.9088071584701538),\n",
       " ('拉脱维亚', 0.8824584484100342),\n",
       " ('意大利', 0.8729462623596191),\n",
       " ('葡萄牙', 0.8672915101051331),\n",
       " ('比利时', 0.8607289791107178),\n",
       " ('匈牙利', 0.8548849821090698),\n",
       " ('丹麦', 0.8534164428710938),\n",
       " ('波兰', 0.8341460227966309),\n",
       " ('奥地利', 0.8164430260658264),\n",
       " ('乌拉圭', 0.8142130374908447),\n",
       " ('斯洛文尼亚', 0.806132972240448),\n",
       " ('爱尔兰', 0.8031853437423706),\n",
       " ('基辅', 0.8002790212631226),\n",
       " ('克罗地亚', 0.8001245260238647),\n",
       " ('立陶宛', 0.7995327115058899),\n",
       " ('津巴布韦', 0.7994288206100464),\n",
       " ('中国香港', 0.7994146347045898),\n",
       " ('白俄罗斯', 0.797614574432373),\n",
       " ('乌克兰', 0.797508180141449),\n",
       " ('保加利亚', 0.7972137331962585)]"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_word2ve.most_similar('捷克', topn=20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\wanga26\\AppData\\Local\\Continuum\\anaconda3\\envs\\cs231n\\lib\\site-packages\\ipykernel\\__main__.py:1: DeprecationWarning: Call to deprecated `most_similar` (Method will be removed in 4.0.0, use self.wv.most_similar() instead).\n",
      "  if __name__ == '__main__':\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[('表示', 0.9060407876968384),\n",
       " ('认为', 0.8647439479827881),\n",
       " ('指出', 0.8475111722946167),\n",
       " ('看来', 0.8290128111839294),\n",
       " ('坦言', 0.8182457089424133),\n",
       " ('介绍', 0.7845182418823242),\n",
       " ('明说', 0.7616503238677979),\n",
       " ('称', 0.7563531398773193),\n",
       " ('文说', 0.7541820406913757),\n",
       " ('透露', 0.7442010641098022),\n",
       " ('告诉', 0.7321453094482422),\n",
       " ('强调', 0.7075918912887573),\n",
       " ('中说', 0.6916170120239258),\n",
       " ('所说', 0.6681145429611206),\n",
       " ('时说', 0.6364998817443848),\n",
       " ('提到', 0.6346604824066162),\n",
       " ('称赞', 0.6289868354797363),\n",
       " ('深有体会', 0.6189056634902954),\n",
       " ('如是说', 0.6172446012496948),\n",
       " ('坦说', 0.6084513068199158),\n",
       " ('道', 0.5804411172866821),\n",
       " ('的话', 0.5803701877593994),\n",
       " ('建议', 0.5786048173904419),\n",
       " ('确信', 0.5748284459114075),\n",
       " ('相信', 0.5684607028961182),\n",
       " ('特别强调', 0.5603676438331604),\n",
       " ('写道', 0.552329957485199),\n",
       " ('普遍认为', 0.5510092377662659),\n",
       " ('直言', 0.5472595691680908),\n",
       " ('说道', 0.5463616847991943),\n",
       " ('地说', 0.5410135984420776),\n",
       " ('叙林', 0.5402573943138123),\n",
       " ('问', 0.5369252562522888),\n",
       " ('林元庆', 0.5299052596092224),\n",
       " ('资深', 0.5293197631835938),\n",
       " ('接受', 0.5281882286071777),\n",
       " ('写信给', 0.5269483923912048),\n",
       " ('请问', 0.5268093347549438),\n",
       " ('解释', 0.521611213684082),\n",
       " ('纳说', 0.5073394775390625),\n",
       " ('呼吁', 0.5058779120445251),\n",
       " ('预言', 0.5038730502128601),\n",
       " ('中称', 0.5026125907897949),\n",
       " ('对瓦姆', 0.5014554262161255),\n",
       " ('提醒', 0.4988548755645752),\n",
       " ('沉得住气', 0.4986202120780945),\n",
       " ('形容', 0.49742016196250916),\n",
       " ('Bill', 0.49705231189727783),\n",
       " ('政务司', 0.4969131648540497),\n",
       " ('来说', 0.4919288158416748),\n",
       " ('非常重视', 0.49119824171066284),\n",
       " ('时称', 0.4864283800125122),\n",
       " ('伟说', 0.4850519001483917),\n",
       " ('眼中', 0.48410722613334656),\n",
       " ('因斯', 0.48403045535087585),\n",
       " ('申明', 0.4837709069252014),\n",
       " ('曼说', 0.48344433307647705),\n",
       " ('知道', 0.48328742384910583),\n",
       " ('兼', 0.48318880796432495),\n",
       " ('国际关系学院', 0.4827888011932373),\n",
       " ('还称', 0.4820823073387146),\n",
       " ('谈到', 0.4818098545074463),\n",
       " ('条法司', 0.4814158082008362),\n",
       " ('声称', 0.4777967929840088),\n",
       " ('托说', 0.4771457612514496),\n",
       " ('这篇大', 0.4767414331436157),\n",
       " ('感谢', 0.47640708088874817),\n",
       " ('估计', 0.4759255647659302),\n",
       " ('深表', 0.47303521633148193),\n",
       " ('并不知道', 0.472874253988266),\n",
       " ('强调指出', 0.47175249457359314),\n",
       " ('尼格', 0.47046181559562683),\n",
       " ('陈说', 0.47009462118148804),\n",
       " ('张伟', 0.4694797694683075),\n",
       " ('三军', 0.46941977739334106),\n",
       " ('及其', 0.4689030647277832),\n",
       " ('首先', 0.4685393273830414),\n",
       " ('八面玲珑', 0.46840739250183105),\n",
       " ('宣称', 0.46801894903182983),\n",
       " ('财政司', 0.46767354011535645),\n",
       " ('副', 0.466545045375824),\n",
       " ('觉得', 0.46451252698898315),\n",
       " ('部副', 0.4643900990486145),\n",
       " ('从零开始', 0.46184730529785156),\n",
       " ('所提', 0.45947176218032837),\n",
       " ('证实', 0.4593220353126526),\n",
       " ('说法', 0.45865845680236816),\n",
       " ('为什么', 0.45851120352745056),\n",
       " ('坚信', 0.4583350121974945),\n",
       " ('一说', 0.4576389491558075),\n",
       " ('所指', 0.4576217532157898),\n",
       " ('格说', 0.45725926756858826),\n",
       " ('供词', 0.4564323425292969),\n",
       " ('与你同在', 0.45613449811935425),\n",
       " ('对此', 0.455534428358078),\n",
       " ('苏珊娜', 0.45548227429389954),\n",
       " ('提及', 0.4554515480995178),\n",
       " ('暗指', 0.45543280243873596),\n",
       " ('而言', 0.4548187851905823),\n",
       " ('嚷嚷', 0.45443472266197205)]"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_word2ve.most_similar('说', topn=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\wanga26\\AppData\\Local\\Continuum\\anaconda3\\envs\\cs231n\\lib\\site-packages\\ipykernel\\__main__.py:1: DeprecationWarning: Call to deprecated `most_similar` (Method will be removed in 4.0.0, use self.wv.most_similar() instead).\n",
      "  if __name__ == '__main__':\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[('指出', 0.9057178497314453),\n",
       " ('表示', 0.8907920718193054),\n",
       " ('说', 0.8647439479827881),\n",
       " ('看来', 0.8213793039321899),\n",
       " ('称', 0.8001088500022888),\n",
       " ('坦言', 0.7940620183944702),\n",
       " ('普遍认为', 0.7686386108398438),\n",
       " ('透露', 0.7454639673233032),\n",
       " ('强调', 0.7233943343162537),\n",
       " ('建议', 0.7016156911849976),\n",
       " ('所说', 0.6716853380203247),\n",
       " ('介绍', 0.6649230718612671),\n",
       " ('相信', 0.6344857215881348),\n",
       " ('告诉', 0.6341235041618347),\n",
       " ('呼吁', 0.6190996766090393),\n",
       " ('提到', 0.6115113496780396),\n",
       " ('不明智', 0.6059680581092834),\n",
       " ('而言', 0.6054043769836426),\n",
       " ('看好', 0.6042644381523132),\n",
       " ('表明', 0.5997722148895264),\n",
       " ('声称', 0.599672794342041),\n",
       " ('事实上', 0.5985932350158691),\n",
       " ('中说', 0.5918971300125122),\n",
       " ('对此', 0.5907052755355835),\n",
       " ('明说', 0.5898700952529907),\n",
       " ('说明', 0.5873211026191711),\n",
       " ('现阶段', 0.586159348487854),\n",
       " ('担心', 0.5804302096366882),\n",
       " ('资深', 0.5792628526687622),\n",
       " ('不过', 0.578912615776062)]"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_word2ve.most_similar('认为', topn=30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\wanga26\\AppData\\Local\\Continuum\\anaconda3\\envs\\cs231n\\lib\\site-packages\\ipykernel\\__main__.py:1: DeprecationWarning: Call to deprecated `most_similar` (Method will be removed in 4.0.0, use self.wv.most_similar() instead).\n",
      "  if __name__ == '__main__':\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[('要求', 0.7343764901161194),\n",
       " ('认为', 0.7016156911849976),\n",
       " ('强调', 0.6876432299613953),\n",
       " ('提醒', 0.6805597543716431),\n",
       " ('指出', 0.6761874556541443),\n",
       " ('慎重', 0.661210298538208),\n",
       " ('必要', 0.6606169939041138),\n",
       " ('解释', 0.660605788230896),\n",
       " ('呼吁', 0.6448093056678772),\n",
       " ('折中', 0.6384769678115845)]"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_word2ve.most_similar('建议', topn=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [],
   "source": [
    "from collections import defaultdict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_related_words(initial_words, model):\n",
    "    \"\"\"\n",
    "    @initial_words are initial words we already know\n",
    "    @model is the word2vec model\n",
    "    \"\"\"\n",
    "    \n",
    "    unseen = initial_words\n",
    "    \n",
    "    seen = defaultdict(int)\n",
    "    \n",
    "    max_size = 500  # could be greater\n",
    "    \n",
    "    while unseen and len(seen) < max_size:\n",
    "        if len(seen) % 50 == 0: \n",
    "            print('seen length : {}'.format(len(seen)))\n",
    "            \n",
    "        node = unseen.pop(0)\n",
    "        \n",
    "        new_expanding = [w for w, s in model.most_similar(node, topn=20)]\n",
    "        \n",
    "        unseen += new_expanding\n",
    "        \n",
    "        seen[node] += 1\n",
    "        \n",
    "        # optimal: 1. score function could be revised\n",
    "        # optimal: 2. using dymanic programming to reduce computing time\n",
    "    \n",
    "    return seen"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_similiar_words(initial_words, model):\n",
    "    max_size = 500\n",
    "    unseen = initial_words\n",
    "    seen = defaultdict(int)\n",
    "    \n",
    "    words_to_explore = initial_words\n",
    "    explored = defaultdict(int)\n",
    "    \n",
    "    while words_to_explore and len(explored) < max_size:\n",
    "        if len(explored) % 10 == 0:\n",
    "            print (len(explored))\n",
    "        word = words_to_explore.pop(0)\n",
    "        similar_words = [w for w, s in model.most_similar(word, topn=20)]\n",
    "        \n",
    "        words_to_explore += similar_words\n",
    "        \n",
    "        explored[word] += 1\n",
    "    \n",
    "    return explored"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "10\n",
      "20\n",
      "30\n",
      "30\n",
      "30\n",
      "30\n",
      "30\n",
      "30\n",
      "30"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\wanga26\\AppData\\Local\\Continuum\\anaconda3\\envs\\cs231n\\lib\\site-packages\\ipykernel\\__main__.py:13: DeprecationWarning: Call to deprecated `most_similar` (Method will be removed in 4.0.0, use self.wv.most_similar() instead).\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "30\n",
      "30\n",
      "30\n",
      "30\n",
      "30\n",
      "30\n",
      "30\n",
      "30\n",
      "30\n",
      "40\n",
      "40\n",
      "50\n",
      "60\n",
      "70\n",
      "80\n",
      "90\n",
      "100\n",
      "110\n",
      "120\n",
      "130\n",
      "140\n",
      "150\n",
      "150\n",
      "150\n",
      "150\n",
      "150\n",
      "150\n",
      "160\n",
      "170\n",
      "170\n",
      "180\n",
      "180\n",
      "190\n",
      "190\n",
      "190\n",
      "200\n",
      "210\n",
      "210\n",
      "220\n",
      "220\n",
      "230\n",
      "240\n",
      "250\n",
      "250\n",
      "260\n",
      "260\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "270\n",
      "280\n",
      "290\n",
      "290\n",
      "300\n",
      "310\n",
      "310\n",
      "320\n",
      "330\n",
      "340\n",
      "350\n",
      "360\n",
      "370\n",
      "380\n",
      "380\n",
      "380\n",
      "390\n",
      "400\n",
      "400\n",
      "400\n",
      "400\n",
      "410\n",
      "420\n",
      "430\n",
      "440\n",
      "440\n",
      "440\n",
      "440\n",
      "440\n",
      "440\n",
      "440\n",
      "440\n",
      "440\n",
      "440\n",
      "440\n",
      "440\n",
      "440\n",
      "440\n",
      "440\n",
      "440\n",
      "440\n",
      "440\n",
      "440\n",
      "440\n",
      "440\n",
      "440\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "450\n",
      "460\n",
      "470\n",
      "480\n",
      "490\n"
     ]
    }
   ],
   "source": [
    "similiar_words = get_similiar_words(['说'], news_word2ve)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 189,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "defaultdict(int,\n",
       "            {'说': 75,\n",
       "             '表示': 75,\n",
       "             '认为': 65,\n",
       "             '指出': 72,\n",
       "             '看来': 55,\n",
       "             '坦言': 74,\n",
       "             '介绍': 59,\n",
       "             '明说': 52,\n",
       "             '称': 51,\n",
       "             '文说': 32,\n",
       "             '透露': 70,\n",
       "             '告诉': 39,\n",
       "             '强调': 46,\n",
       "             '中说': 36,\n",
       "             '所说': 52,\n",
       "             '时说': 23,\n",
       "             '提到': 57,\n",
       "             '称赞': 36,\n",
       "             '深有体会': 19,\n",
       "             '如是说': 20,\n",
       "             '坦说': 18,\n",
       "             '普遍认为': 32,\n",
       "             '呼吁': 26,\n",
       "             '建议': 26,\n",
       "             '接受': 13,\n",
       "             '相信': 15,\n",
       "             '不明智': 6,\n",
       "             '而言': 14,\n",
       "             '看好': 9,\n",
       "             '表明': 6,\n",
       "             '特别强调': 29,\n",
       "             '条法司': 6,\n",
       "             '中称': 21,\n",
       "             '来说': 8,\n",
       "             '直言': 22,\n",
       "             '谈到': 17,\n",
       "             '确信': 15,\n",
       "             '觉得': 12,\n",
       "             '地说': 12,\n",
       "             '说道': 21,\n",
       "             '写信给': 6,\n",
       "             '获悉': 6,\n",
       "             '了解': 6,\n",
       "             '据介绍': 6,\n",
       "             '中山大学': 6,\n",
       "             '市政管理': 6,\n",
       "             '部副': 6,\n",
       "             '卫': 6,\n",
       "             '供图': 6,\n",
       "             '农艺师': 10,\n",
       "             '武说': 6,\n",
       "             '郭': 6,\n",
       "             '祁': 10,\n",
       "             '连生': 6,\n",
       "             '黄进': 8,\n",
       "             '东升': 6,\n",
       "             '三军': 11,\n",
       "             '陆': 6,\n",
       "             '八面玲珑': 6,\n",
       "             '彩霞': 6,\n",
       "             '保国': 6,\n",
       "             '国利': 6,\n",
       "             '海涛': 6,\n",
       "             '还称': 16,\n",
       "             '宣称': 8,\n",
       "             '说法': 8,\n",
       "             '提及': 20,\n",
       "             '声称': 17,\n",
       "             '知情': 12,\n",
       "             '并称': 5,\n",
       "             '否认': 5,\n",
       "             '回应': 5,\n",
       "             '证实': 13,\n",
       "             '匿名': 6,\n",
       "             '因斯': 4,\n",
       "             '科说': 4,\n",
       "             '农办': 4,\n",
       "             '张晓明': 6,\n",
       "             '高松': 6,\n",
       "             '韩长': 5,\n",
       "             '国际关系学院': 6,\n",
       "             '第七任': 4,\n",
       "             '副': 4,\n",
       "             '政务司': 7,\n",
       "             '常务副': 4,\n",
       "             '利维': 4,\n",
       "             '国家航天局': 4,\n",
       "             '给出': 14,\n",
       "             '问': 5,\n",
       "             '见到': 5,\n",
       "             '看到': 5,\n",
       "             '先对': 5,\n",
       "             '聊起': 5,\n",
       "             '洛登': 5,\n",
       "             '鲁金博': 5,\n",
       "             '黄勇贤摄': 5,\n",
       "             '小春': 5,\n",
       "             '留意到': 5,\n",
       "             '每经': 5,\n",
       "             '眼中': 9,\n",
       "             '苏津': 5,\n",
       "             '重申': 8,\n",
       "             '一贯': 5,\n",
       "             '提出': 5,\n",
       "             '阐述': 11,\n",
       "             '高度重视': 5,\n",
       "             '坚持': 5,\n",
       "             '明确提出': 5,\n",
       "             '方针': 5,\n",
       "             '所作': 5,\n",
       "             '遵循': 5,\n",
       "             '敦促': 8,\n",
       "             '反复强调': 8,\n",
       "             '时称': 9,\n",
       "             '上称': 6,\n",
       "             '写道': 14,\n",
       "             '置评': 4,\n",
       "             '评述': 3,\n",
       "             '努钦称': 3,\n",
       "             '援引': 4,\n",
       "             '里亚布': 3,\n",
       "             '公开批评': 4,\n",
       "             '问道': 5,\n",
       "             '引用': 11,\n",
       "             '形容': 6,\n",
       "             '所指': 5,\n",
       "             '正如': 7,\n",
       "             '意思': 5,\n",
       "             '反驳': 6,\n",
       "             '解释': 14,\n",
       "             '承认': 5,\n",
       "             '回答': 11,\n",
       "             '预言': 5,\n",
       "             '谈论': 16,\n",
       "             '此行': 5,\n",
       "             '不久前': 3,\n",
       "             '他称': 3,\n",
       "             '盛赞': 7,\n",
       "             '联访': 3,\n",
       "             '新闻频道': 3,\n",
       "             '导报': 3,\n",
       "             '期待': 5,\n",
       "             '就此': 3,\n",
       "             '谈及': 7,\n",
       "             '森访': 3,\n",
       "             '列举': 6,\n",
       "             '知道': 6,\n",
       "             '言': 13,\n",
       "             '说明': 7,\n",
       "             '描述': 6,\n",
       "             '感谢': 10,\n",
       "             '自豪': 3,\n",
       "             '赞美': 3,\n",
       "             '高度赞扬': 3,\n",
       "             '讽刺': 3,\n",
       "             '惊叹': 3,\n",
       "             '批判': 3,\n",
       "             '非常感谢': 5,\n",
       "             '感慨': 9,\n",
       "             '历史课': 3,\n",
       "             '感激': 4,\n",
       "             '由衷': 6,\n",
       "             '喜出望外': 3,\n",
       "             '问起': 2,\n",
       "             '翃': 2,\n",
       "             '佩服': 2,\n",
       "             '深有感触': 2,\n",
       "             '灏': 2,\n",
       "             '干劲十足': 2,\n",
       "             '薰': 2,\n",
       "             '说实话': 6,\n",
       "             '拭目以待': 3,\n",
       "             '大夫': 2,\n",
       "             '魏': 2,\n",
       "             '下定决心': 2,\n",
       "             '自称为': 2,\n",
       "             '福兴': 1,\n",
       "             '做客': 1,\n",
       "             '副总干事': 1,\n",
       "             '德说': 1,\n",
       "             '库说': 3,\n",
       "             '琳': 1,\n",
       "             '小林': 1,\n",
       "             '培根': 1,\n",
       "             '伊戈尔': 1,\n",
       "             '新文化': 1,\n",
       "             '国务院台湾事务办公室': 1,\n",
       "             '灿': 1,\n",
       "             '迈克': 1,\n",
       "             '特别感谢': 2,\n",
       "             '苏珊娜': 2,\n",
       "             '公使衔': 2,\n",
       "             '兼': 2,\n",
       "             '延斯': 2,\n",
       "             '高级顾问': 2,\n",
       "             '事务部长': 2,\n",
       "             '中国人民对外友好协会': 2,\n",
       "             '参赞': 2,\n",
       "             '布鲁金斯': 2,\n",
       "             '罗伯特': 2,\n",
       "             '交流中心': 2,\n",
       "             '奥尔布赖特': 2,\n",
       "             '阿泽': 2,\n",
       "             '贸易部': 2,\n",
       "             '康斯坦丁': 2,\n",
       "             '关注': 4,\n",
       "             '猜测': 3,\n",
       "             '令': 3,\n",
       "             '质疑': 3,\n",
       "             '暗示': 3,\n",
       "             '鉴于': 3,\n",
       "             '对此': 3,\n",
       "             '抨击': 3,\n",
       "             '批评': 6,\n",
       "             '电影界': 3,\n",
       "             '热议': 3,\n",
       "             '针对': 3,\n",
       "             '动向': 4,\n",
       "             '资深': 3,\n",
       "             '主张': 3,\n",
       "             '反对': 3,\n",
       "             '抵制': 3,\n",
       "             '指责': 3,\n",
       "             '对朝': 3,\n",
       "             '谴责': 3,\n",
       "             '促使': 3,\n",
       "             '采取措施': 3,\n",
       "             '应': 6,\n",
       "             '挑衅': 3,\n",
       "             '敌对': 3,\n",
       "             '申明': 3,\n",
       "             '公敌': 3,\n",
       "             '施压': 3,\n",
       "             '要求': 3,\n",
       "             '提醒': 3,\n",
       "             '慎重': 3,\n",
       "             '必要': 3,\n",
       "             '折中': 3,\n",
       "             '意见建议': 3,\n",
       "             '为此': 3,\n",
       "             '采纳': 3,\n",
       "             '大包大揽': 3,\n",
       "             '专家建议': 3,\n",
       "             '还应': 3,\n",
       "             '鼓励': 3,\n",
       "             '相应': 3,\n",
       "             '产假': 1,\n",
       "             '获知': 1,\n",
       "             '召集': 2,\n",
       "             '检视': 1,\n",
       "             '上海理工大学': 1,\n",
       "             '后勤部门': 1,\n",
       "             '驻巴': 1,\n",
       "             '得知': 1,\n",
       "             '伊曼': 1,\n",
       "             '介入': 1,\n",
       "             '取得联系': 1,\n",
       "             '希望': 2,\n",
       "             '期望': 2,\n",
       "             '坚信': 3,\n",
       "             '当然': 7,\n",
       "             '将来': 2,\n",
       "             '或许': 4,\n",
       "             '肯定': 2,\n",
       "             '让': 2,\n",
       "             '非常高兴': 4,\n",
       "             '理解': 2,\n",
       "             '指望': 3,\n",
       "             '恐怕': 2,\n",
       "             '前途': 2,\n",
       "             '当头棒喝': 1,\n",
       "             '众所周知': 1,\n",
       "             '微妙': 1,\n",
       "             '视而不见': 1,\n",
       "             '过分': 1,\n",
       "             '误读': 1,\n",
       "             '正面': 1,\n",
       "             '骨牌效应': 1,\n",
       "             '显而易见': 2,\n",
       "             '究其原因': 1,\n",
       "             '虚幻': 1,\n",
       "             '何种': 1,\n",
       "             '意图': 1,\n",
       "             '难免': 1,\n",
       "             '疑虑': 1,\n",
       "             '不难看出': 1,\n",
       "             '何来': 1,\n",
       "             '根源': 1,\n",
       "             '歧视性': 1,\n",
       "             '理性': 1,\n",
       "             '来讲': 3,\n",
       "             '其实': 3,\n",
       "             '但是': 4,\n",
       "             '事实上': 4,\n",
       "             '还是': 3,\n",
       "             '实际上': 2,\n",
       "             '本身': 3,\n",
       "             '显然': 3,\n",
       "             '所以': 3,\n",
       "             '感兴趣': 3,\n",
       "             '短浅': 3,\n",
       "             '而且': 3,\n",
       "             '不过': 3,\n",
       "             '一方面': 2,\n",
       "             '来看': 2,\n",
       "             '衡量': 3,\n",
       "             '这方面': 2,\n",
       "             '依赖于': 1,\n",
       "             '看重': 1,\n",
       "             '前景': 1,\n",
       "             '强劲': 1,\n",
       "             '航空业': 1,\n",
       "             '水涨船高': 1,\n",
       "             '蓬勃发展': 1,\n",
       "             '受益': 1,\n",
       "             '乐观': 1,\n",
       "             '发展前景': 1,\n",
       "             '现阶段': 2,\n",
       "             '可观': 1,\n",
       "             '充满信心': 3,\n",
       "             '增长速度': 1,\n",
       "             '令人信服': 1,\n",
       "             '非常明显': 1,\n",
       "             '指向': 1,\n",
       "             '迄今为止': 1,\n",
       "             '调门': 1,\n",
       "             '释放': 1,\n",
       "             '尽管': 1,\n",
       "             '印证': 1,\n",
       "             '依赖': 1,\n",
       "             '反映': 2,\n",
       "             '总体而言': 1,\n",
       "             '但': 2,\n",
       "             '敏感': 1,\n",
       "             '利欲熏心': 1,\n",
       "             '潜在': 1,\n",
       "             '远远不够': 3,\n",
       "             '领会': 3,\n",
       "             '谈谈': 3,\n",
       "             '高瞻远瞩': 3,\n",
       "             '非常重视': 3,\n",
       "             '十分重视': 3,\n",
       "             '理': 3,\n",
       "             '积极行动': 3,\n",
       "             '认真学习': 3,\n",
       "             '直言不讳': 3,\n",
       "             '谋略': 3,\n",
       "             '请问': 3,\n",
       "             '责无旁贷': 3,\n",
       "             '司长': 1,\n",
       "             '司': 1,\n",
       "             '副司长': 1,\n",
       "             '国家外国专家局': 1,\n",
       "             '信息部': 1,\n",
       "             '管理司': 1,\n",
       "             '财政司': 1,\n",
       "             '局局长': 1,\n",
       "             '赋': 1,\n",
       "             '部长': 1,\n",
       "             '事务局': 1,\n",
       "             '范一飞': 1,\n",
       "             '署': 1,\n",
       "             '钟山': 1,\n",
       "             '投资司': 1,\n",
       "             '研究会': 1,\n",
       "             '局长': 1,\n",
       "             '详解': 1,\n",
       "             '揭秘': 1,\n",
       "             '登出': 1,\n",
       "             '伊原': 1,\n",
       "             '澄清': 1,\n",
       "             '属实': 1,\n",
       "             '日媒': 1,\n",
       "             '沃尔夫': 1,\n",
       "             '中央日报': 1,\n",
       "             '哈里斯': 1,\n",
       "             '毕竟': 2,\n",
       "             '很大': 1,\n",
       "             '深感': 4,\n",
       "             '怼': 3,\n",
       "             '感叹': 5,\n",
       "             '坦承': 3,\n",
       "             '更令': 2,\n",
       "             '感到遗憾': 2,\n",
       "             '要说': 3,\n",
       "             '乌纱帽': 2,\n",
       "             '出乎': 2,\n",
       "             '感到高兴': 2,\n",
       "             '回顾': 2,\n",
       "             '赞赏': 2,\n",
       "             '借此机会': 2,\n",
       "             '有加': 2,\n",
       "             '李义东': 2,\n",
       "             '极好': 2,\n",
       "             '庆幸': 2,\n",
       "             '意料': 1,\n",
       "             '只不过': 1,\n",
       "             '也许': 2,\n",
       "             '注定': 1,\n",
       "             '取悦': 1,\n",
       "             '置身事外': 1,\n",
       "             '为什么': 1,\n",
       "             '更何况': 1,\n",
       "             '走运': 1,\n",
       "             '明白': 1,\n",
       "             '本来': 1,\n",
       "             '可是': 1,\n",
       "             '就让': 1,\n",
       "             '做些': 1,\n",
       "             '从没': 1,\n",
       "             '想不到': 1,\n",
       "             '羡慕': 1,\n",
       "             '心里': 2,\n",
       "             '从来不': 1,\n",
       "             '挺': 1,\n",
       "             '懂': 1,\n",
       "             '是不是': 1,\n",
       "             '而已': 1,\n",
       "             '溢于言表': 1,\n",
       "             '眼里': 1,\n",
       "             '说起': 1,\n",
       "             '谈起': 1,\n",
       "             '那位': 2,\n",
       "             '口中': 2,\n",
       "             '总是': 1,\n",
       "             '倍感': 1,\n",
       "             '说完': 2,\n",
       "             '回忆起': 2,\n",
       "             '无比': 1,\n",
       "             '想起': 2,\n",
       "             '不已': 1,\n",
       "             '忍不住': 1,\n",
       "             '开玩笑': 1,\n",
       "             '骂': 1,\n",
       "             '记得': 1,\n",
       "             '哽咽': 1,\n",
       "             '听闻': 1,\n",
       "             '离去': 1,\n",
       "             '一脸': 1,\n",
       "             '大发雷霆': 1,\n",
       "             '拜访': 1,\n",
       "             '写给': 1,\n",
       "             '催': 1,\n",
       "             '自称': 1,\n",
       "             '请来': 1,\n",
       "             '昭惠': 1,\n",
       "             '曾任': 1,\n",
       "             '痛批': 1,\n",
       "             '女婿': 1,\n",
       "             '过世': 1,\n",
       "             '会同': 1,\n",
       "             '广东省': 1,\n",
       "             '交管局': 1,\n",
       "             '厦门市': 1,\n",
       "             '刑侦局': 1,\n",
       "             '市卫': 1,\n",
       "             '濮阳市': 1,\n",
       "             '科技厅': 1,\n",
       "             '民政厅': 1,\n",
       "             '广州市': 1,\n",
       "             '旅游委': 1,\n",
       "             '湖南省': 1,\n",
       "             'OECD': 1,\n",
       "             '交通委': 1,\n",
       "             '杭州市': 1,\n",
       "             '公安局': 1,\n",
       "             '下发': 1,\n",
       "             '工商局': 1,\n",
       "             '教育厅': 1,\n",
       "             '认识': 1,\n",
       "             '注意': 1,\n",
       "             '测算': 1,\n",
       "             '中来': 1,\n",
       "             '统计': 1,\n",
       "             '港官': 1,\n",
       "             '感受': 1,\n",
       "             '精神鼓舞': 1,\n",
       "             '参考报': 1,\n",
       "             '切身感受': 1,\n",
       "             '体会': 1,\n",
       "             '解答': 1,\n",
       "             '估算': 1,\n",
       "             '肖娟': 1,\n",
       "             '目前为止': 1,\n",
       "             '查阅': 1,\n",
       "             '梳理': 1,\n",
       "             '据悉': 1,\n",
       "             '一期': 1,\n",
       "             '宣传周': 1,\n",
       "             '推广应用': 1,\n",
       "             '据统计': 1,\n",
       "             '首批': 1,\n",
       "             '光谷': 1,\n",
       "             '三期': 1,\n",
       "             '招商': 1,\n",
       "             '普查': 1,\n",
       "             '专项': 1,\n",
       "             '博会': 1,\n",
       "             '法院系统': 1,\n",
       "             '农牧厅': 1})"
      ]
     },
     "execution_count": 189,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "similiar_words"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('说', 75),\n",
       " ('表示', 75),\n",
       " ('坦言', 74),\n",
       " ('指出', 72),\n",
       " ('透露', 70),\n",
       " ('认为', 65),\n",
       " ('介绍', 59),\n",
       " ('提到', 57),\n",
       " ('看来', 55),\n",
       " ('明说', 52),\n",
       " ('所说', 52),\n",
       " ('称', 51),\n",
       " ('强调', 46),\n",
       " ('告诉', 39),\n",
       " ('中说', 36),\n",
       " ('称赞', 36),\n",
       " ('文说', 32),\n",
       " ('普遍认为', 32),\n",
       " ('特别强调', 29),\n",
       " ('呼吁', 26),\n",
       " ('建议', 26),\n",
       " ('时说', 23),\n",
       " ('直言', 22),\n",
       " ('中称', 21),\n",
       " ('说道', 21),\n",
       " ('如是说', 20),\n",
       " ('提及', 20),\n",
       " ('深有体会', 19),\n",
       " ('坦说', 18),\n",
       " ('谈到', 17),\n",
       " ('声称', 17),\n",
       " ('还称', 16),\n",
       " ('谈论', 16),\n",
       " ('相信', 15),\n",
       " ('确信', 15),\n",
       " ('而言', 14),\n",
       " ('给出', 14),\n",
       " ('写道', 14),\n",
       " ('解释', 14),\n",
       " ('接受', 13),\n",
       " ('证实', 13),\n",
       " ('言', 13),\n",
       " ('觉得', 12),\n",
       " ('地说', 12),\n",
       " ('知情', 12),\n",
       " ('三军', 11),\n",
       " ('阐述', 11),\n",
       " ('引用', 11),\n",
       " ('回答', 11),\n",
       " ('农艺师', 10),\n",
       " ('祁', 10),\n",
       " ('感谢', 10),\n",
       " ('看好', 9),\n",
       " ('眼中', 9),\n",
       " ('时称', 9),\n",
       " ('感慨', 9),\n",
       " ('来说', 8),\n",
       " ('黄进', 8),\n",
       " ('宣称', 8),\n",
       " ('说法', 8),\n",
       " ('重申', 8),\n",
       " ('敦促', 8),\n",
       " ('反复强调', 8),\n",
       " ('政务司', 7),\n",
       " ('正如', 7),\n",
       " ('盛赞', 7),\n",
       " ('谈及', 7),\n",
       " ('说明', 7),\n",
       " ('当然', 7),\n",
       " ('不明智', 6),\n",
       " ('表明', 6),\n",
       " ('条法司', 6),\n",
       " ('写信给', 6),\n",
       " ('获悉', 6),\n",
       " ('了解', 6),\n",
       " ('据介绍', 6),\n",
       " ('中山大学', 6),\n",
       " ('市政管理', 6),\n",
       " ('部副', 6),\n",
       " ('卫', 6),\n",
       " ('供图', 6),\n",
       " ('武说', 6),\n",
       " ('郭', 6),\n",
       " ('连生', 6),\n",
       " ('东升', 6),\n",
       " ('陆', 6),\n",
       " ('八面玲珑', 6),\n",
       " ('彩霞', 6),\n",
       " ('保国', 6),\n",
       " ('国利', 6),\n",
       " ('海涛', 6),\n",
       " ('匿名', 6),\n",
       " ('张晓明', 6),\n",
       " ('高松', 6),\n",
       " ('国际关系学院', 6),\n",
       " ('上称', 6),\n",
       " ('形容', 6),\n",
       " ('反驳', 6),\n",
       " ('列举', 6),\n",
       " ('知道', 6),\n",
       " ('描述', 6),\n",
       " ('由衷', 6),\n",
       " ('说实话', 6),\n",
       " ('批评', 6),\n",
       " ('应', 6),\n",
       " ('并称', 5),\n",
       " ('否认', 5),\n",
       " ('回应', 5),\n",
       " ('韩长', 5),\n",
       " ('问', 5),\n",
       " ('见到', 5),\n",
       " ('看到', 5),\n",
       " ('先对', 5),\n",
       " ('聊起', 5),\n",
       " ('洛登', 5),\n",
       " ('鲁金博', 5),\n",
       " ('黄勇贤摄', 5),\n",
       " ('小春', 5),\n",
       " ('留意到', 5),\n",
       " ('每经', 5),\n",
       " ('苏津', 5),\n",
       " ('一贯', 5),\n",
       " ('提出', 5),\n",
       " ('高度重视', 5),\n",
       " ('坚持', 5),\n",
       " ('明确提出', 5),\n",
       " ('方针', 5),\n",
       " ('所作', 5),\n",
       " ('遵循', 5),\n",
       " ('问道', 5),\n",
       " ('所指', 5),\n",
       " ('意思', 5),\n",
       " ('承认', 5),\n",
       " ('预言', 5),\n",
       " ('此行', 5),\n",
       " ('期待', 5),\n",
       " ('非常感谢', 5),\n",
       " ('感叹', 5),\n",
       " ('因斯', 4),\n",
       " ('科说', 4),\n",
       " ('农办', 4),\n",
       " ('第七任', 4),\n",
       " ('副', 4),\n",
       " ('常务副', 4),\n",
       " ('利维', 4),\n",
       " ('国家航天局', 4),\n",
       " ('置评', 4),\n",
       " ('援引', 4),\n",
       " ('公开批评', 4),\n",
       " ('感激', 4),\n",
       " ('关注', 4),\n",
       " ('动向', 4),\n",
       " ('或许', 4),\n",
       " ('非常高兴', 4),\n",
       " ('但是', 4),\n",
       " ('事实上', 4),\n",
       " ('深感', 4),\n",
       " ('评述', 3),\n",
       " ('努钦称', 3),\n",
       " ('里亚布', 3),\n",
       " ('不久前', 3),\n",
       " ('他称', 3),\n",
       " ('联访', 3),\n",
       " ('新闻频道', 3),\n",
       " ('导报', 3),\n",
       " ('就此', 3),\n",
       " ('森访', 3),\n",
       " ('自豪', 3),\n",
       " ('赞美', 3),\n",
       " ('高度赞扬', 3),\n",
       " ('讽刺', 3),\n",
       " ('惊叹', 3),\n",
       " ('批判', 3),\n",
       " ('历史课', 3),\n",
       " ('喜出望外', 3),\n",
       " ('拭目以待', 3),\n",
       " ('库说', 3),\n",
       " ('猜测', 3),\n",
       " ('令', 3),\n",
       " ('质疑', 3),\n",
       " ('暗示', 3),\n",
       " ('鉴于', 3),\n",
       " ('对此', 3),\n",
       " ('抨击', 3),\n",
       " ('电影界', 3),\n",
       " ('热议', 3),\n",
       " ('针对', 3),\n",
       " ('资深', 3),\n",
       " ('主张', 3),\n",
       " ('反对', 3),\n",
       " ('抵制', 3),\n",
       " ('指责', 3),\n",
       " ('对朝', 3),\n",
       " ('谴责', 3),\n",
       " ('促使', 3),\n",
       " ('采取措施', 3),\n",
       " ('挑衅', 3),\n",
       " ('敌对', 3),\n",
       " ('申明', 3),\n",
       " ('公敌', 3),\n",
       " ('施压', 3),\n",
       " ('要求', 3),\n",
       " ('提醒', 3),\n",
       " ('慎重', 3),\n",
       " ('必要', 3),\n",
       " ('折中', 3),\n",
       " ('意见建议', 3),\n",
       " ('为此', 3),\n",
       " ('采纳', 3),\n",
       " ('大包大揽', 3),\n",
       " ('专家建议', 3),\n",
       " ('还应', 3),\n",
       " ('鼓励', 3),\n",
       " ('相应', 3),\n",
       " ('坚信', 3),\n",
       " ('指望', 3),\n",
       " ('来讲', 3),\n",
       " ('其实', 3),\n",
       " ('还是', 3),\n",
       " ('本身', 3),\n",
       " ('显然', 3),\n",
       " ('所以', 3),\n",
       " ('感兴趣', 3),\n",
       " ('短浅', 3),\n",
       " ('而且', 3),\n",
       " ('不过', 3),\n",
       " ('衡量', 3),\n",
       " ('充满信心', 3),\n",
       " ('远远不够', 3),\n",
       " ('领会', 3),\n",
       " ('谈谈', 3),\n",
       " ('高瞻远瞩', 3),\n",
       " ('非常重视', 3),\n",
       " ('十分重视', 3),\n",
       " ('理', 3),\n",
       " ('积极行动', 3),\n",
       " ('认真学习', 3),\n",
       " ('直言不讳', 3),\n",
       " ('谋略', 3),\n",
       " ('请问', 3),\n",
       " ('责无旁贷', 3),\n",
       " ('怼', 3),\n",
       " ('坦承', 3),\n",
       " ('要说', 3),\n",
       " ('问起', 2),\n",
       " ('翃', 2),\n",
       " ('佩服', 2),\n",
       " ('深有感触', 2),\n",
       " ('灏', 2),\n",
       " ('干劲十足', 2),\n",
       " ('薰', 2),\n",
       " ('大夫', 2),\n",
       " ('魏', 2),\n",
       " ('下定决心', 2),\n",
       " ('自称为', 2),\n",
       " ('特别感谢', 2),\n",
       " ('苏珊娜', 2),\n",
       " ('公使衔', 2),\n",
       " ('兼', 2),\n",
       " ('延斯', 2),\n",
       " ('高级顾问', 2),\n",
       " ('事务部长', 2),\n",
       " ('中国人民对外友好协会', 2),\n",
       " ('参赞', 2),\n",
       " ('布鲁金斯', 2),\n",
       " ('罗伯特', 2),\n",
       " ('交流中心', 2),\n",
       " ('奥尔布赖特', 2),\n",
       " ('阿泽', 2),\n",
       " ('贸易部', 2),\n",
       " ('康斯坦丁', 2),\n",
       " ('召集', 2),\n",
       " ('希望', 2),\n",
       " ('期望', 2),\n",
       " ('将来', 2),\n",
       " ('肯定', 2),\n",
       " ('让', 2),\n",
       " ('理解', 2),\n",
       " ('恐怕', 2),\n",
       " ('前途', 2),\n",
       " ('显而易见', 2),\n",
       " ('实际上', 2),\n",
       " ('一方面', 2),\n",
       " ('来看', 2),\n",
       " ('这方面', 2),\n",
       " ('现阶段', 2),\n",
       " ('反映', 2),\n",
       " ('但', 2),\n",
       " ('毕竟', 2),\n",
       " ('更令', 2),\n",
       " ('感到遗憾', 2),\n",
       " ('乌纱帽', 2),\n",
       " ('出乎', 2),\n",
       " ('感到高兴', 2),\n",
       " ('回顾', 2),\n",
       " ('赞赏', 2),\n",
       " ('借此机会', 2),\n",
       " ('有加', 2),\n",
       " ('李义东', 2),\n",
       " ('极好', 2),\n",
       " ('庆幸', 2),\n",
       " ('也许', 2),\n",
       " ('心里', 2),\n",
       " ('那位', 2),\n",
       " ('口中', 2),\n",
       " ('说完', 2),\n",
       " ('回忆起', 2),\n",
       " ('想起', 2),\n",
       " ('福兴', 1),\n",
       " ('做客', 1),\n",
       " ('副总干事', 1),\n",
       " ('德说', 1),\n",
       " ('琳', 1),\n",
       " ('小林', 1),\n",
       " ('培根', 1),\n",
       " ('伊戈尔', 1),\n",
       " ('新文化', 1),\n",
       " ('国务院台湾事务办公室', 1),\n",
       " ('灿', 1),\n",
       " ('迈克', 1),\n",
       " ('产假', 1),\n",
       " ('获知', 1),\n",
       " ('检视', 1),\n",
       " ('上海理工大学', 1),\n",
       " ('后勤部门', 1),\n",
       " ('驻巴', 1),\n",
       " ('得知', 1),\n",
       " ('伊曼', 1),\n",
       " ('介入', 1),\n",
       " ('取得联系', 1),\n",
       " ('当头棒喝', 1),\n",
       " ('众所周知', 1),\n",
       " ('微妙', 1),\n",
       " ('视而不见', 1),\n",
       " ('过分', 1),\n",
       " ('误读', 1),\n",
       " ('正面', 1),\n",
       " ('骨牌效应', 1),\n",
       " ('究其原因', 1),\n",
       " ('虚幻', 1),\n",
       " ('何种', 1),\n",
       " ('意图', 1),\n",
       " ('难免', 1),\n",
       " ('疑虑', 1),\n",
       " ('不难看出', 1),\n",
       " ('何来', 1),\n",
       " ('根源', 1),\n",
       " ('歧视性', 1),\n",
       " ('理性', 1),\n",
       " ('依赖于', 1),\n",
       " ('看重', 1),\n",
       " ('前景', 1),\n",
       " ('强劲', 1),\n",
       " ('航空业', 1),\n",
       " ('水涨船高', 1),\n",
       " ('蓬勃发展', 1),\n",
       " ('受益', 1),\n",
       " ('乐观', 1),\n",
       " ('发展前景', 1),\n",
       " ('可观', 1),\n",
       " ('增长速度', 1),\n",
       " ('令人信服', 1),\n",
       " ('非常明显', 1),\n",
       " ('指向', 1),\n",
       " ('迄今为止', 1),\n",
       " ('调门', 1),\n",
       " ('释放', 1),\n",
       " ('尽管', 1),\n",
       " ('印证', 1),\n",
       " ('依赖', 1),\n",
       " ('总体而言', 1),\n",
       " ('敏感', 1),\n",
       " ('利欲熏心', 1),\n",
       " ('潜在', 1),\n",
       " ('司长', 1),\n",
       " ('司', 1),\n",
       " ('副司长', 1),\n",
       " ('国家外国专家局', 1),\n",
       " ('信息部', 1),\n",
       " ('管理司', 1),\n",
       " ('财政司', 1),\n",
       " ('局局长', 1),\n",
       " ('赋', 1),\n",
       " ('部长', 1),\n",
       " ('事务局', 1),\n",
       " ('范一飞', 1),\n",
       " ('署', 1),\n",
       " ('钟山', 1),\n",
       " ('投资司', 1),\n",
       " ('研究会', 1),\n",
       " ('局长', 1),\n",
       " ('详解', 1),\n",
       " ('揭秘', 1),\n",
       " ('登出', 1),\n",
       " ('伊原', 1),\n",
       " ('澄清', 1),\n",
       " ('属实', 1),\n",
       " ('日媒', 1),\n",
       " ('沃尔夫', 1),\n",
       " ('中央日报', 1),\n",
       " ('哈里斯', 1),\n",
       " ('很大', 1),\n",
       " ('意料', 1),\n",
       " ('只不过', 1),\n",
       " ('注定', 1),\n",
       " ('取悦', 1),\n",
       " ('置身事外', 1),\n",
       " ('为什么', 1),\n",
       " ('更何况', 1),\n",
       " ('走运', 1),\n",
       " ('明白', 1),\n",
       " ('本来', 1),\n",
       " ('可是', 1),\n",
       " ('就让', 1),\n",
       " ('做些', 1),\n",
       " ('从没', 1),\n",
       " ('想不到', 1),\n",
       " ('羡慕', 1),\n",
       " ('从来不', 1),\n",
       " ('挺', 1),\n",
       " ('懂', 1),\n",
       " ('是不是', 1),\n",
       " ('而已', 1),\n",
       " ('溢于言表', 1),\n",
       " ('眼里', 1),\n",
       " ('说起', 1),\n",
       " ('谈起', 1),\n",
       " ('总是', 1),\n",
       " ('倍感', 1),\n",
       " ('无比', 1),\n",
       " ('不已', 1),\n",
       " ('忍不住', 1),\n",
       " ('开玩笑', 1),\n",
       " ('骂', 1),\n",
       " ('记得', 1),\n",
       " ('哽咽', 1),\n",
       " ('听闻', 1),\n",
       " ('离去', 1),\n",
       " ('一脸', 1),\n",
       " ('大发雷霆', 1),\n",
       " ('拜访', 1),\n",
       " ('写给', 1),\n",
       " ('催', 1),\n",
       " ('自称', 1),\n",
       " ('请来', 1),\n",
       " ('昭惠', 1),\n",
       " ('曾任', 1),\n",
       " ('痛批', 1),\n",
       " ('女婿', 1),\n",
       " ('过世', 1),\n",
       " ('会同', 1),\n",
       " ('广东省', 1),\n",
       " ('交管局', 1),\n",
       " ('厦门市', 1),\n",
       " ('刑侦局', 1),\n",
       " ('市卫', 1),\n",
       " ('濮阳市', 1),\n",
       " ('科技厅', 1),\n",
       " ('民政厅', 1),\n",
       " ('广州市', 1),\n",
       " ('旅游委', 1),\n",
       " ('湖南省', 1),\n",
       " ('OECD', 1),\n",
       " ('交通委', 1),\n",
       " ('杭州市', 1),\n",
       " ('公安局', 1),\n",
       " ('下发', 1),\n",
       " ('工商局', 1),\n",
       " ('教育厅', 1),\n",
       " ('认识', 1),\n",
       " ('注意', 1),\n",
       " ('测算', 1),\n",
       " ('中来', 1),\n",
       " ('统计', 1),\n",
       " ('港官', 1),\n",
       " ('感受', 1),\n",
       " ('精神鼓舞', 1),\n",
       " ('参考报', 1),\n",
       " ('切身感受', 1),\n",
       " ('体会', 1),\n",
       " ('解答', 1),\n",
       " ('估算', 1),\n",
       " ('肖娟', 1),\n",
       " ('目前为止', 1),\n",
       " ('查阅', 1),\n",
       " ('梳理', 1),\n",
       " ('据悉', 1),\n",
       " ('一期', 1),\n",
       " ('宣传周', 1),\n",
       " ('推广应用', 1),\n",
       " ('据统计', 1),\n",
       " ('首批', 1),\n",
       " ('光谷', 1),\n",
       " ('三期', 1),\n",
       " ('招商', 1),\n",
       " ('普查', 1),\n",
       " ('专项', 1),\n",
       " ('博会', 1),\n",
       " ('法院系统', 1),\n",
       " ('农牧厅', 1)]"
      ]
     },
     "execution_count": 190,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sorted(similiar_words.items(), key=lambda x: x[1], reverse=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "97927"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(news_word2ve.wv.vocab)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "seen length : 0\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\wanga26\\AppData\\Local\\Continuum\\anaconda3\\envs\\cs231n\\lib\\site-packages\\ipykernel\\__main__.py:19: DeprecationWarning: Call to deprecated `most_similar` (Method will be removed in 4.0.0, use self.wv.most_similar() instead).\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "seen length : 50\n",
      "seen length : 100\n",
      "seen length : 150\n",
      "seen length : 150\n",
      "seen length : 150\n",
      "seen length : 150\n",
      "seen length : 150\n",
      "seen length : 150\n",
      "seen length : 200\n",
      "seen length : 250\n",
      "seen length : 250\n",
      "seen length : 300\n",
      "seen length : 350\n",
      "seen length : 400\n",
      "seen length : 400\n",
      "seen length : 400\n",
      "seen length : 400\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n",
      "seen length : 450\n"
     ]
    }
   ],
   "source": [
    "related_words = get_related_words(['说', '表示'], news_word2ve)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('表示', 92),\n",
       " ('说', 89),\n",
       " ('指出', 86),\n",
       " ('坦言', 85),\n",
       " ('透露', 84),\n",
       " ('认为', 77),\n",
       " ('介绍', 69),\n",
       " ('提到', 65),\n",
       " ('看来', 64),\n",
       " ('明说', 60),\n",
       " ('称', 60),\n",
       " ('所说', 59),\n",
       " ('强调', 54),\n",
       " ('告诉', 46),\n",
       " ('中说', 43),\n",
       " ('称赞', 40),\n",
       " ('文说', 37),\n",
       " ('普遍认为', 37),\n",
       " ('特别强调', 34),\n",
       " ('呼吁', 31),\n",
       " ('建议', 31),\n",
       " ('时说', 28),\n",
       " ('中称', 25),\n",
       " ('直言', 24),\n",
       " ('如是说', 23),\n",
       " ('提及', 23),\n",
       " ('深有体会', 22),\n",
       " ('说道', 22),\n",
       " ('坦说', 21),\n",
       " ('谈到', 20),\n",
       " ('还称', 19),\n",
       " ('声称', 19),\n",
       " ('相信', 17),\n",
       " ('谈论', 17),\n",
       " ('而言', 16),\n",
       " ('确信', 16),\n",
       " ('证实', 16),\n",
       " ('给出', 16),\n",
       " ('写道', 16),\n",
       " ('解释', 16)]"
      ]
     },
     "execution_count": 87,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sorted(related_words.items(), key=lambda x: x[1], reverse=True)[:40]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\wanga26\\AppData\\Local\\Continuum\\anaconda3\\envs\\cs231n\\lib\\site-packages\\ipykernel\\__main__.py:1: DeprecationWarning: Call to deprecated `most_similar` (Method will be removed in 4.0.0, use self.wv.most_similar() instead).\n",
      "  if __name__ == '__main__':\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[('表示', 0.9060407876968384),\n",
       " ('认为', 0.8647439479827881),\n",
       " ('指出', 0.8475111722946167),\n",
       " ('看来', 0.8290128111839294),\n",
       " ('坦言', 0.8182457089424133),\n",
       " ('介绍', 0.7845182418823242),\n",
       " ('明说', 0.7616503238677979),\n",
       " ('称', 0.7563531398773193),\n",
       " ('文说', 0.7541820406913757),\n",
       " ('透露', 0.7442010641098022),\n",
       " ('告诉', 0.7321453094482422),\n",
       " ('强调', 0.7075918912887573),\n",
       " ('中说', 0.6916170120239258),\n",
       " ('所说', 0.6681145429611206),\n",
       " ('时说', 0.6364998817443848),\n",
       " ('提到', 0.6346604824066162),\n",
       " ('称赞', 0.6289868354797363),\n",
       " ('深有体会', 0.6189056634902954),\n",
       " ('如是说', 0.6172446012496948),\n",
       " ('坦说', 0.6084513068199158),\n",
       " ('道', 0.5804411172866821),\n",
       " ('的话', 0.5803701877593994),\n",
       " ('建议', 0.5786048173904419),\n",
       " ('确信', 0.5748284459114075),\n",
       " ('相信', 0.5684607028961182),\n",
       " ('特别强调', 0.5603676438331604),\n",
       " ('写道', 0.552329957485199),\n",
       " ('普遍认为', 0.5510092377662659),\n",
       " ('直言', 0.5472595691680908),\n",
       " ('说道', 0.5463616847991943),\n",
       " ('地说', 0.5410135984420776),\n",
       " ('叙林', 0.5402573943138123),\n",
       " ('问', 0.5369252562522888),\n",
       " ('林元庆', 0.5299052596092224),\n",
       " ('资深', 0.5293197631835938),\n",
       " ('接受', 0.5281882286071777),\n",
       " ('写信给', 0.5269483923912048),\n",
       " ('请问', 0.5268093347549438),\n",
       " ('解释', 0.521611213684082),\n",
       " ('纳说', 0.5073394775390625)]"
      ]
     },
     "execution_count": 88,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_word2ve.most_similar('说', topn=40)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## TFIDF Key words"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'此外   自 本周   6 月 12 日   起   除 小米 手机 6 等 15 款 机型 外   其余 机型 已 暂停 更新 发布   含 开发 版   体验版 内测   稳定版 暂不受 影响   以 确保 工程师 可以 集中 全部 精力 进行 系统优化 工作   有人 猜测 这 也 是 将 精力 主要 用到 MIUI   9 的 研发 之中   MIUI   8 去年 5 月 发布   距今已有 一年 有余   也 是 时候 更新换代 了   当然   关于 MIUI   9 的 确切 信息   我们 还是 等待 官方消息'"
      ]
     },
     "execution_count": 89,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_content[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [],
   "source": [
    "def document_frequency(word): \n",
    "    return sum(1 for n in news_content if word in n)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "70342"
      ]
     },
     "execution_count": 91,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "document_frequency('的')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [],
   "source": [
    "import math"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [],
   "source": [
    "def idf(word):\n",
    "    \"\"\"Gets the inversed document frequency\"\"\"\n",
    "    return math.log10(len(news_content) / document_frequency(word))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "idf('的') < idf('小米')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 98,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.6736077228396447"
      ]
     },
     "execution_count": 98,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "idf('华为')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [],
   "source": [
    "def tf(word, document):\n",
    "    \"\"\"\n",
    "    Gets the term frequemcy of a @word in a @document.\n",
    "    \"\"\"\n",
    "    words = document.split()\n",
    "    \n",
    "    return sum(1 for w in words if w == word)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'?\\r\\n\\u3000\\u3000巨丰早评：市场将再次探底\\r\\n\\u3000\\u3000【巨丰观点】\\r\\n\\u3000\\u3000周四大盘冲高回落，2500余只个股下跌。上证50、沪深300指数顽强收红。早盘大盘低开高走，展开反弹，沪指冲破120日均线压力位；午后大盘跳水，个股普跌。盘面上，保险、电力、银行、民航机场、券商、房地产、人工智能、迪士尼、网络安全等板块涨幅居前。雄安新区、福州新区、医疗器械、丝绸之路等跌幅居前。\\r\\n\\u3000\\u3000金融股全线大涨：保险板块涨3%：新华保险、中国太保、中国人寿涨4%，中国平安涨2%；西水股份、天茂集团盘中大涨7%以上。银行板块涨2%：招商银行大涨8%，贵阳银行、兴业银行、浦发银行、兴业银行等涨幅居前。金融股是A股纳入MSCI指数的受益者，后市可以逢低关注。\\r\\n\\u3000\\u3000盘中万达电影、复星医药等白马股闪崩，导致午后市场出现大幅跳水。盘后消息，银监会于6月中旬要求各家银行排查包括万达、海航集团、复星、浙江罗森内里在内数家企业的授信及风险分析，排查对象多是近年来海外投资比较凶猛、在银行业敞口较大的民营企业集团。\\r\\n\\u3000\\u3000巨丰投顾认为自5月以来，大盘已经走出了3重底的走势，并开启中级反弹行情。沪指在上证50及白马股推动下率先反弹，但遭遇60日、120日均线强压力而出现调整。MSCI纳入A股后，沪深300指数创17个月新高，但2500余股下跌，市场分化明显。随着6月末市场流动性紧张的可能出现，市场将再次出现探底。建议投资者重点关注半年报大幅预增的错杀股和具备估值优势的大蓝筹。\\r\\n\\u3000\\u3000天信投资：尾盘跳水原因揭秘 伏击的品种曝光\\r\\n\\u3000\\u3000周四市场点评：\\r\\n\\u3000\\u3000沪深两市早盘纷纷低开，但是开盘之后权重板块崛起，其中保险和银行表现尤其明显，低矮动股指快速拉升，并顺势突破了60日均线和半年线的束缚；在主板拉升的过程中，市场的分化格局也较明显，题材板块和创业板表现相对较弱。不过虽然市场在早盘的上攻，但是成交量萎缩明教明显，所以这种上涨比较虚。果不其然，午后个股再度纷纷杀跌，同时股指也是顺势回落，沪市的半年线得而复失，形成长长的上影线；而创业板股指在回落的过程中失守5日均线的支撑。整体上周四市场全天保持宽幅震荡走势，尤其是沪市冲高回落。保险、银行、中字头、举牌等品种涨幅居前；钛金属、区块链、雄安新区、民营医院等品种领跌。\\r\\n\\u3000\\u3000今日市场预测：\\r\\n\\u3000\\u3000在周三收市的评论中，我们指出周四市场会出现阶段性的普涨行情，而周四早盘三大股指齐齐上涨，即使创业板股指涨幅较小，但是市场确实也给我们呈现的普涨格局。但是午后的跳水，包括创业板股指先下跌，随后沪指尾盘的快速下杀，使得市场从普涨瞬间到普跌，市场的大起大落来的非常刺激。从三大股指周四呈现给大家的K线组合来看，预计今日市场齐跌的概率较大，不过创业板合格题材板块或具备一定的抗跌性。\\r\\n\\u3000\\u3000短期行情判断：\\r\\n\\u3000\\u3000从大的环境上来看，市场目前本身不具备大面积和大空间的反弹基础，因为目前无论是从宏观面、货币基本面或者从国际经济和政治的角度来看，都不具备这样的条件，所以反应到市场中来，只能是结构性、局部性的投机性机会。而最近半个月以来，市场的走势也确实符合局部性、结构性投机的走势。\\r\\n\\u3000\\u3000周四市场沪指形成带长长上影线的倒锤阴线；创业板股指形成的是跌破5日和10日均线的中阴线，这种类型的K线形态，几乎都是市场开始调整的先兆，所以短期市场而言，悲观的情绪或升温。\\r\\n\\u3000\\u3000后市投资建议：\\r\\n\\u3000\\u3000短期建议投资者一定要控制好自身的仓位。中期继续看好题材板块的崛起，特别是次新股、高送转、重组股等等可以不断的去伏击，依旧具备超跌反弹的性质。\\r\\n'"
      ]
     },
     "execution_count": 100,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "content['content'][11]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "6"
      ]
     },
     "execution_count": 101,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tf('银行', news_content[11])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "6"
      ]
     },
     "execution_count": 102,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tf('创业板', news_content[11])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.506757120518876"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "idf('创业板')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.4550169427748936"
      ]
     },
     "execution_count": 104,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "idf('银行')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.9923664854640575"
      ]
     },
     "execution_count": 105,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "idf('短期')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tf('短期', news_content[11])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_keywords_of_a_ducment(document):\n",
    "    words = set(document.split())\n",
    "    \n",
    "    tfidf = [\n",
    "        (w, tf(w, document) * idf(w)) for w in words\n",
    "    ]\n",
    "    \n",
    "    tfidf = sorted(tfidf, key=lambda x: x[1], reverse=True)\n",
    "    \n",
    "    return tfidf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'此外   自 本周   6 月 12 日   起   除 小米 手机 6 等 15 款 机型 外   其余 机型 已 暂停 更新 发布   含 开发 版   体验版 内测   稳定版 暂不受 影响   以 确保 工程师 可以 集中 全部 精力 进行 系统优化 工作   有人 猜测 这 也 是 将 精力 主要 用到 MIUI   9 的 研发 之中   MIUI   8 去年 5 月 发布   距今已有 一年 有余   也 是 时候 更新换代 了   当然   关于 MIUI   9 的 确切 信息   我们 还是 等待 官方消息'"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_content[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'巨丰 早评   市场 将 再次 探底   巨丰 观点   周四 大盘 冲高 回落   2500 余 只 个股 下跌   上证 50   沪 深 300 指数 顽强 收红   早盘 大盘 低开 高 走   展开 反弹   沪 指 冲破 120 日 均线 压力 位   午后 大盘 跳水   个股 普跌   盘面 上   保险   电力   银行   民航机场   券商   房地产   人工智能   迪士尼   网络安全 等 板块 涨幅 居前   雄安 新区   福州 新区   医疗器械   丝绸之路 等 跌幅 居前   金融股 全线 大涨   保险 板块 涨 3   新华 保险   中国 太保   中国 人寿 涨 4   中国 平安 涨 2   西水股份   天茂 集团 盘中 大涨 7   以上   银行 板块 涨 2   招商银行 大涨 8   贵阳 银行   兴业银行   浦发银行   兴业银行 等 涨幅 居前   金融股 是 A股 纳入 MSCI 指数 的 受益者   后市 可以 逢 低 关注   盘中 万达 电影   复星 医药 等 白马股 闪崩   导致 午后 市场 出现 大幅 跳水   盘后 消息   银监会 于 6 月 中旬 要求 各家 银行 排查 包括 万达   海航 集团   复星   浙江 罗森 内里 在内 数家 企业 的 授信 及 风险 分析   排查 对象 多 是 近年来 海外投资 比较 凶猛   在 银行业 敞口 较大 的 民营企业 集团   巨丰 投顾 认为 自 5 月 以来   大盘 已经 走出 了 3 重底 的 走势   并 开启 中级 反弹 行情   沪 指 在 上证 50 及 白马股 推动 下 率先 反弹   但 遭遇 60 日   120 日 均线 强 压力 而 出现 调整   MSCI 纳入 A股 后   沪 深 300 指数 创 17 个 月 新高   但 2500 余股 下跌   市场 分化 明显   随着 6 月末 市场 流动性 紧张 的 可能 出现   市场 将 再次出现 探底   建议 投资者 重点 关注 半年报 大幅 预增 的 错 杀 股 和 具备 估值 优势 的 大 蓝筹   天信 投资   尾盘 跳水 原因 揭秘   伏击 的 品种 曝光   周四 市场 点评   沪 深 两市 早盘 纷纷 低开   但是 开盘 之后 权重 板块 崛起   其中 保险 和 银行 表现 尤其 明显   低矮 动 股指 快速 拉升   并 顺势 突破 了 60 日 均线 和 半年线 的 束缚   在 主板 拉升 的 过程 中   市场 的 分化 格局 也 较 明显   题材 板块 和 创业板 表现 相对 较弱   不过 虽然 市场 在 早盘 的 上攻   但是 成交量 萎缩 明教 明显   所以 这种 上涨 比较 虚   果不其然   午后 个股 再度 纷纷 杀跌   同时 股指 也 是 顺势 回落   沪市 的 半年线 得而复失   形成 长长的 上影线   而 创业板 股指 在 回落 的 过程 中 失守 5 日 均线 的 支撑   整体 上周四 市场 全天 保持 宽幅 震荡 走势   尤其 是 沪市 冲高 回落   保险   银行   中 字头   举牌 等 品种 涨幅 居前   钛 金属   区块 链   雄安 新区   民营 医院 等 品种 领跌   今日 市场预测   在 周三 收市 的 评论 中   我们 指出 周四 市场 会 出现 阶段性 的 普涨 行情   而 周四 早盘 三大 股指 齐齐 上涨   即使 创业板 股指 涨幅 较 小   但是 市场 确实 也 给 我们 呈现 的 普涨 格局   但是 午后 的 跳水   包括 创业板 股指 先 下跌   随后 沪 指 尾盘 的 快速 下 杀   使得 市场 从 普涨 瞬间 到 普跌   市场 的 大起大落 来 的 非常 刺激   从三大 股指 周四 呈现 给 大家 的 K 线 组合 来看   预计 今日 市场 齐跌 的 概率 较大   不过 创业板 合格 题材 板块 或 具备 一定 的 抗跌性   短期 行情 判断   从 大 的 环境 上 来看   市场 目前 本身 不 具备 大面积 和 大 空间 的 反弹 基础   因为 目前 无论是 从 宏观 面   货币 基本面 或者 从 国际 经济 和 政治 的 角度 来看   都 不 具备 这样 的 条件   所以 反应 到 市场 中来   只能 是 结构性   局部性 的 投机性 机会   而 最近 半个 月 以来   市场 的 走势 也 确实 符合 局部性   结构性 投机 的 走势   周四 市场 沪 指 形成 带 长长 上影线 的 倒 锤 阴线   创业板 股指 形成 的 是 跌破 5 日 和 10 日 均线 的 中 阴线   这种 类型 的 K 线 形态   几乎 都 是 市场 开始 调整 的 先兆   所以 短期 市场 而言   悲观 的 情绪 或 升温   后市 投资 建议   短期 建议 投资者 一定 要 控制 好 自身 的 仓位   中期 继续 看好 题材 板块 的 崛起   特别 是 次新股   高送 转   重组股 等等 可以 不断 的 去 伏击   依旧 具备 超跌 反弹 的 性质'"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_content[11]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " "
     ]
    }
   ],
   "source": [
    "%prun get_keywords_of_a_ducment(news_content[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [],
   "source": [
    "machine_new_keywords = get_keywords_of_a_ducment(news_content[101])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'中国 经济网 6 月 23 日讯   记者 刘瑾   日前   中国 机械 工业 百强 企业   汽车 三十 强 企业名单 出炉   在 由 中国 机械 工业 联合会   中国汽车工业协会 主办 的   中国 机械 工业 百强 企业   汽车 工业 三十 强 企业 信息 发布会   上   中国 机械 工业 联合会 会长 王瑞祥 指出   入围 企业 总体 呈现 了 持续增长 态势   2016 年 总 规模   平均 规模 均 创新 高   我国 机械 工业 总体 上 呈现 稳中 向 好 的 发展 态势   主要 经济指标 出现 恢复性 增长   同时   王瑞祥 也 表示   机械 工业 长期 积累 的 结构性 矛盾 还 没有 根本性 改变   新 的 发展 业态   发展 模式 远 未 形成   保持 平稳 健康 发展 的 基础 还 不 牢固   产业 升级   新旧 动能 转换   实现 由大 变强 任重道远   相关 数据 显示   2016 年 机械 百强 入围 企业 规模 是 15   4 亿元   比 上年 略 高   总 规模 达到 16992 亿元   平均 规模 169   92 亿元   两项 均 创新 高   相比 前 两年 的 前十强 排名   名单 基本 没有 什么 变化   只是 工程 机械行业 的 徐工   三一 和 中联重科 由于 之前 市场 环境 等 原因   整体 排位 后移 了 一 两位   但 仍 全部 在 十强 之 列   但 考虑 到 工程 机械行业 最近 以来 的 强烈 增长 态势   明年 该 行业 整体 恢复性 可能 较大   中机联 执行 副会长 陈斌 表示   机械 百强 的 年 换位 率 2005 年 曾 高达 32   此后 逐年 下降   2012 年降 至 6   不过 近几年 又 在 上升   2016 年 升至 22   说明 市场 动荡   不 确定 因素 增加   根据 中机联 公布 的 数据   2017 年 1 至 5 月 机械 增加值 增速 好于 去年 但 呈现 逐月 回落 的 态势   比如 1 至 5 月 增速 为 10   7   比 1 至 4 月 回落 0   2 点   此外 利润总额 也 出现 了 同样 的 走势   专家 分析   这是 由于 钢铁 等 原材料 价格 近期 出现 较大 增幅   使得 工业 企业 成本 压力 增大   产量 同比 增长 表现 抢眼 的 是 工程 机械类 产品   已经 有 8 个 月 保持高速 增长   同时   陈斌 也 表示 了 担忧   比如 价格指数 缓慢 回升 但远 低于 工业   原材料 指数   固定资产 投资 增速 仍 较 低迷   增速 在 多年 回落 之后 继续 在 低位 震荡 徘徊   他 表示   行业 经济运行 仍 面临 一定 的 下行 压力   机械行业 主要 服务 的 钢铁   煤炭   电力 及 石油   化工 等 行业 普遍 处于 产业结构 深度 调整期   其 装备 需求 短期 难以 大幅 增长   还有 经过 多年 的 高速 发展   各类 机械 产品 的 社会 保有量 均 达到 了 相当规模   对 现役 设备 的 更新改造 维护 已 成为 需求 中 的 重要 部分   这 增加 了 增量 回升 的 难度'"
      ]
     },
     "execution_count": 112,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_content[101]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('机械', 19.73268423154147),\n",
       " ('工业', 13.726799287798574),\n",
       " ('百强', 12.218936930008129),\n",
       " ('机械行业', 11.732905915902746),\n",
       " ('王瑞祥', 9.904722647584947),\n",
       " ('中机联', 9.904722647584947),\n",
       " ('增速', 8.057368905265053),\n",
       " ('态势', 7.971371787277902),\n",
       " ('恢复性', 6.867694767829172),\n",
       " ('规模', 6.861746056315247),\n",
       " ('回落', 6.643121992377128),\n",
       " ('增长', 6.604587773896507),\n",
       " ('企业', 6.260042817487905),\n",
       " ('陈斌', 5.9311791790524575),\n",
       " ('呈现', 5.279381611367784),\n",
       " ('原材料', 5.215938100214726),\n",
       " ('16992', 4.9523613237924735),\n",
       " ('但远', 4.9523613237924735),\n",
       " ('刘瑾', 4.651331328128492),\n",
       " ('三十', 4.6438668975349),\n",
       " ('入围', 4.629743188559925),\n",
       " ('钢铁', 4.588699854270723),\n",
       " ('回升', 4.553165964236777),\n",
       " ('联合会', 4.4415451172114695),\n",
       " ('行业', 4.256149034981132),\n",
       " ('前十强', 4.253391319456455),\n",
       " ('后移', 4.253391319456455),\n",
       " ('发展', 4.24347205022841),\n",
       " ('亿元', 4.194707226838383),\n",
       " ('更新改造', 4.17421007340883),\n",
       " ('机械类', 4.17421007340883),\n",
       " ('相当规模', 4.17421007340883),\n",
       " ('年降', 4.107263283778217),\n",
       " ('徐工', 3.9981188143531488),\n",
       " ('工程', 3.9916862993694178),\n",
       " ('出现', 3.7560113879094406),\n",
       " ('中联重科', 3.748241341136549),\n",
       " ('仍', 3.665948565614492),\n",
       " ('较大', 3.65832649752095),\n",
       " ('中国汽车工业协会', 3.6513313281284923),\n",
       " ('变强', 3.6099386429702673),\n",
       " ('总体', 3.591018845444925),\n",
       " ('保持高速', 3.5373879758216558),\n",
       " ('换位', 3.5052032924502545),\n",
       " ('比如', 3.4905315667465113),\n",
       " ('逐月', 3.4752400690728114),\n",
       " ('由大', 3.4752400690728114),\n",
       " ('表示', 3.4536423582977496),\n",
       " ('企业名单', 3.4472113454725677),\n",
       " ('均', 3.413941587967606),\n",
       " ('十强', 3.3960588230251862),\n",
       " ('三一', 3.3503013324645115),\n",
       " ('平均', 3.342655913089492),\n",
       " ('整体', 3.322429124121176),\n",
       " ('调整期', 3.318892868212887),\n",
       " ('压力', 3.305708050183972),\n",
       " ('经济指标', 3.21199863429823),\n",
       " ('汽车', 3.195737446406074),\n",
       " ('保有量', 3.188933330229536),\n",
       " ('经济网', 3.126286521091647),\n",
       " ('利润总额', 3.071547731511682),\n",
       " ('抢眼', 3.007878651642305),\n",
       " ('任重道远', 3.0029713171475607),\n",
       " ('多年', 2.9986978653420366),\n",
       " ('根本性', 2.9885734964469184),\n",
       " ('排位', 2.970090090752905),\n",
       " ('169', 2.948039950009831),\n",
       " ('中国', 2.917491209470893),\n",
       " ('徘徊', 2.9109686386342486),\n",
       " ('新旧', 2.8954564724560012),\n",
       " ('现役', 2.8916634834388617),\n",
       " ('经济运行', 2.884175462046312),\n",
       " ('需求', 2.8632317045459827),\n",
       " ('好于', 2.8519907786749106),\n",
       " ('固定资产', 2.806233288114236),\n",
       " ('至', 2.784114021467203),\n",
       " ('持续增长', 2.762029625622182),\n",
       " ('了', 2.7530014092101593),\n",
       " ('增加值', 2.7375174757447756),\n",
       " ('达到', 2.7333500906799535),\n",
       " ('出炉', 2.7068486559783236),\n",
       " ('增加', 2.665284116361493),\n",
       " ('产业结构', 2.653508247382767),\n",
       " ('副会长', 2.6448652858792605),\n",
       " ('由于', 2.624157388191586),\n",
       " ('增大', 2.617907572641543),\n",
       " ('低位', 2.5925258414525856),\n",
       " ('动荡', 2.5685459578120424),\n",
       " ('创新', 2.5681372266885054),\n",
       " ('增量', 2.5614262166890946),\n",
       " ('牢固', 2.556161976696737),\n",
       " ('2016', 2.5450574183796917),\n",
       " ('价格指数', 2.5258500624278986),\n",
       " ('产品', 2.5209690024662845),\n",
       " ('2005', 2.513028629962211),\n",
       " ('上年', 2.492968836033243),\n",
       " ('逐年', 2.4914634810359257),\n",
       " ('缓慢', 2.488468334806566),\n",
       " ('两项', 2.46948774018372),\n",
       " ('但', 2.4593444283982095),\n",
       " ('稳中', 2.4445054520966427),\n",
       " ('转换', 2.4312232400884373),\n",
       " ('数据', 2.4284304701608277),\n",
       " ('低迷', 2.4273165167556283),\n",
       " ('业态', 2.4021329707373793),\n",
       " ('近几年', 2.362411722466766),\n",
       " ('化工', 2.3546661378669613),\n",
       " ('下行', 2.3503013324645115),\n",
       " ('日讯', 2.3385195019164042),\n",
       " ('增幅', 2.3364112721360724),\n",
       " ('煤炭', 2.3301473008261784),\n",
       " ('升至', 2.318892868212887),\n",
       " ('震荡', 2.304001312811542),\n",
       " ('动能', 2.2849083709025195),\n",
       " ('明年', 2.2473533644591375),\n",
       " ('走势', 2.221579048126084),\n",
       " ('主要', 2.2196289745949174),\n",
       " ('会长', 2.1670314887817064),\n",
       " ('高达', 2.1627806116280484),\n",
       " ('担忧', 2.1585709391016548),\n",
       " ('市场', 2.1353584391728972),\n",
       " ('没有', 2.1292630530088865),\n",
       " ('两位', 2.112255229335716),\n",
       " ('的', 2.102932231028948),\n",
       " ('积累', 2.0992717939406083),\n",
       " ('产量', 2.0962368795501733),\n",
       " ('电力', 2.092622757595327),\n",
       " ('矛盾', 2.09202331722148),\n",
       " ('结构性', 2.0860739847082788),\n",
       " ('石油', 2.0647440234567376),\n",
       " ('平稳', 2.0647440234567376),\n",
       " ('也', 2.035789095863084),\n",
       " ('同时', 2.0271830020187123),\n",
       " ('难度', 2.022431763707886),\n",
       " ('此后', 1.9942754752713885),\n",
       " ('92', 1.9942754752713885),\n",
       " ('短期', 1.9923664854640575),\n",
       " ('使得', 1.968735036667939),\n",
       " ('2012', 1.9655895895262288),\n",
       " ('名单', 1.9655895895262288),\n",
       " ('低于', 1.949763343072565),\n",
       " ('强烈', 1.9074297776433136),\n",
       " ('深度', 1.9027557111975004),\n",
       " ('说明', 1.8996673818675058),\n",
       " ('排名', 1.8663576181740917),\n",
       " ('高速', 1.8628094409060196),\n",
       " ('装备', 1.833105434514537),\n",
       " ('难以', 1.8062332881142356),\n",
       " ('普遍', 1.8031422111370938),\n",
       " ('指数', 1.7780107263130935),\n",
       " ('上升', 1.7662529439792682),\n",
       " ('最近', 1.7496006363992738),\n",
       " ('主办', 1.744188797125352),\n",
       " ('同样', 1.7333009913436124),\n",
       " ('高', 1.7282840637671162),\n",
       " ('同比', 1.7265353323305803),\n",
       " ('各类', 1.7183505062052942),\n",
       " ('下降', 1.70783681222239),\n",
       " ('大幅', 1.7051247742857096),\n",
       " ('处于', 1.6934846944203423),\n",
       " ('因素', 1.6915746551374973),\n",
       " ('维护', 1.6670290961485892),\n",
       " ('相比', 1.6643357884041108),\n",
       " ('两年', 1.6603268877977373),\n",
       " ('只是', 1.6500303951080744),\n",
       " ('考虑', 1.6489492531957317),\n",
       " ('32', 1.636600833126739),\n",
       " ('等', 1.6323131825512693),\n",
       " ('近期', 1.6282847440529873),\n",
       " ('之前', 1.6169034231030894),\n",
       " ('确定', 1.615101784042198),\n",
       " ('成本', 1.6021133054583108),\n",
       " ('改变', 1.6013384712083498),\n",
       " ('设备', 1.6007582513733447),\n",
       " ('表现', 1.5861252000741806),\n",
       " ('发布会', 1.5825890351955108),\n",
       " ('升级', 1.5788976021601047),\n",
       " ('面临', 1.5593128573756954),\n",
       " ('执行', 1.5315807042439082),\n",
       " ('价格', 1.5182716853745828),\n",
       " ('变化', 1.51556281356067),\n",
       " ('原因', 1.4901131084374761),\n",
       " ('一定', 1.489215187066124),\n",
       " ('长期', 1.4777080704304109),\n",
       " ('还', 1.4730246953620385),\n",
       " ('强', 1.4717574528800272),\n",
       " ('模式', 1.471062050459618),\n",
       " ('日前', 1.4650815073494048),\n",
       " ('什么', 1.4632469544135542),\n",
       " ('保持', 1.4515761508750178),\n",
       " ('公布', 1.4443705989727822),\n",
       " ('不过', 1.4431588014613708),\n",
       " ('健康', 1.436355093406426),\n",
       " ('全部', 1.4222897549550955),\n",
       " ('基本', 1.4194796043850764),\n",
       " ('经过', 1.4090384231455615),\n",
       " ('之后', 1.4028676105774605),\n",
       " ('此外', 1.3910213823335724),\n",
       " ('指出', 1.389711651580557),\n",
       " ('分析', 1.388050413831871),\n",
       " ('还有', 1.375559427963561),\n",
       " ('形成', 1.3654366156476532),\n",
       " ('专家', 1.3543372514582839),\n",
       " ('我国', 1.3365164409177714),\n",
       " ('总', 1.327598250797306),\n",
       " ('去年', 1.2942544882860807),\n",
       " ('产业', 1.2508796881715465),\n",
       " ('投资', 1.2463266630781231),\n",
       " ('信息', 1.2461557819105031),\n",
       " ('略', 1.2382831588106176),\n",
       " ('继续', 1.2350237410686098),\n",
       " ('基础', 1.2270215078827365),\n",
       " ('根据', 1.2232775667488618),\n",
       " ('环境', 1.2142946090150044),\n",
       " ('显示', 1.1870653940944171),\n",
       " ('曾', 1.1781364189235546),\n",
       " ('又', 1.1611620772935982),\n",
       " ('远', 1.157664196900475),\n",
       " ('22', 1.1569679888611843),\n",
       " ('可能', 1.1455032942736563),\n",
       " ('以来', 1.1343995192604797),\n",
       " ('部分', 1.121644374355576),\n",
       " ('23', 1.1108141585359204),\n",
       " ('实现', 1.1061479599130863),\n",
       " ('服务', 1.0882657503682263),\n",
       " ('率', 1.0870204181678897),\n",
       " ('社会', 1.0746445229274968),\n",
       " ('已经', 1.0650253933933065),\n",
       " ('相关', 1.059599089176657),\n",
       " ('较', 1.0157973186572078),\n",
       " ('不', 1.0069391592825077),\n",
       " ('列', 0.9869714535773513),\n",
       " ('这是', 0.9835516098796154),\n",
       " ('成为', 0.9830852283035417),\n",
       " ('重要', 0.9685096048010019),\n",
       " ('15', 0.9544509564457302),\n",
       " ('该', 0.9133043057590299),\n",
       " ('未', 0.8997827987478156),\n",
       " ('比', 0.8420969845527588),\n",
       " ('是', 0.8138857280827148),\n",
       " ('10', 0.7960848277894573),\n",
       " ('6', 0.7622602583437212),\n",
       " ('由', 0.7609103093275782),\n",
       " ('向', 0.7486215151586154),\n",
       " ('好', 0.7288653828300791),\n",
       " ('点', 0.7228074598113371),\n",
       " ('他', 0.7074907017272314),\n",
       " ('上', 0.7073166822180775),\n",
       " ('及', 0.6828250310365068),\n",
       " ('已', 0.6788276436411759),\n",
       " ('其', 0.6604599233270064),\n",
       " ('之', 0.5990414795497642),\n",
       " ('8', 0.5787505954795428),\n",
       " ('4', 0.5700969769950187),\n",
       " ('5', 0.5511525867465363),\n",
       " ('到', 0.5374380896146497),\n",
       " ('个', 0.5035468301229501),\n",
       " ('这', 0.5000074536188772),\n",
       " ('对', 0.49485789721916834),\n",
       " ('前', 0.4735784562858743),\n",
       " ('有', 0.44307813505580745),\n",
       " ('为', 0.38692675348471445),\n",
       " ('和', 0.3810835072902879),\n",
       " ('年', 0.3785594723181429),\n",
       " ('1', 0.3482445783310308),\n",
       " ('记者', 0.30906660129308033),\n",
       " ('在', 0.2748809271090362),\n",
       " ('一', 0.23694300119741732),\n",
       " ('2017', 0.2022542388651758),\n",
       " ('月', 0.17435784429986328),\n",
       " ('中', 0.17187119251677352),\n",
       " ('7', 0.15297252207708978),\n",
       " ('0', 0.12495973836320823),\n",
       " ('2', 0.12008239257713019),\n",
       " ('新', 0.031997949117575065)]"
      ]
     },
     "execution_count": 113,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_keywords_of_a_ducment(news_content[101])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Wordcloud"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "import wordcloud"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [],
   "source": [
    "wc = wordcloud.WordCloud('./data/SourceHanSerifSC-Regular.otf')\n",
    "# we could download the font from https://github.com/Computing-Intelligence/datasource"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'原 标题   44 岁 女子 跑 深圳 约会 网友 被 拒   暴雨 中 裸身 奔走   深圳 交警 微博称   昨日 清晨 交警 发现 有 一 女子 赤裸 上身   行走 在 南坪 快速 上   期间 还 起 了 轻生 年头   一辅警 发现 后 赶紧 为 其 披上 黄衣   并 一路 劝说 她   那么 事发 时   到底 都 发生 了 些 什么 呢   南 都 记者 带 您 一起 还原 现场   南 都 记者 在 龙岗 大队 坂田 中队 见到 了 辅警 刘青   发现 女生 的 辅警   一位 外表 高大 帅气   说话 略带 些 腼腆 的 90 后 青年   刘青 介绍   6 月 16 日 早上 7 时 36 分   他 正在 环城 南路 附近 值勤   接到 中队 关于 一位 女子 裸身 进入 机动车 可能 有 危险 的 警情   随后 骑着 小 铁骑 开始 沿路 寻找   大概 花 了 十多分钟 在 南坪 大道 坂田 出口 往 龙岗 方向 的 逆行 辅道 上 发现 该 女子   女子 身上 一丝不挂 地逆 车流 而行   时 走时 停   时 坐 时 躺   险象环生   刘青停 好 小 铁骑   和 另外 一名 巡防 员 追 了 上去   发现 女子 的 情绪 很 低落   话 不 多   刘青 尝试 和 女子 交流   劝说 女子 离开   可 女子 并 不 愿意 接受   继续 缓慢 地往 南坪 快速路 的 主干道 上 走 去   此时 路边 上 已经 聚集 了 很 市民 围观   为了 不 刺激 女子 的 情绪   刘青 和 巡防 员 一边 盯 着 女子 一边 驱赶 着 围观 的 群众   现场 还原   从 警方 提供 的 一份 视频 了解 到   16 日 早上 7 时 25 分   女子 出现 在 坂雪岗 大道 与 环城 南路 的 监控 视频 中   此时 女子 还 穿着 白色 的 内裤   正 沿着 坂雪岗 大道 往 南坪 快速 的 方向 缓慢 地 走 着   当时 正值 上班 高峰期   十字路口 的 车流 已经 排起 了 长队   当 女子 出现 时   路上 的 市民 纷纷 驻足 观望   不少 车辆 也 放慢 了 速度   但 女子 并 不 为 市民 观望 停下 脚步   依然 缓慢 走 着   当 女子 行进 到 十字路口 中间 时   一辆 大 货车 挡住 了 镜头   但是 当 女子 再次出现 镜头 时   可以 发现 女子 已经 没 穿 内裤 了   全身 裸露 继续 朝着 南坪 快速 方向 走 去   记者 发现   视频 中 女子 周围 并 没有 人 尾随 或者 上前 劝止 的 市民   一 大清早 路上 看到 这样 的 情况   恐怕 大家 都 没 办法 淡定   面对 这一 情况   刘青 表示   一 开始 根本 不敢 看 她   心里 挺 别扭   感觉 很 尴尬   但 当 刘青 跟随 女子 上 了 南坪 快速路 主干道 时   女子 作出 了 让 人 意想不到 的 举动   她 突然 靠近 护栏 要 从 上面 跳下去   刘青 赶忙 冲上去 拉住 了 女子 的 手   将 其 控制 住 并 远离 护栏   碍于 女子 没有 穿衣服   刘青递 上 衣服   女子 没 接受 还 把 衣服 扔 到 排水沟 里   继续 往前走   没 办法 刘青 只能 紧紧 拉 着 她 的 一只 手 跟 在 后面   刘青 一路上 耐心 地 开导 安慰 她   但 只 听到 她 不断 地 重复 着 一句 话   要是 你 也 遭遇 我 的 事   你 也 会 这样 的   期间 她 还 不时 试图 挣脱 刘青 的 手要 冲向 护栏 往下跳   就 这样   我 被 牵 着 走 了 大概 十多分钟   天 突然 下起 了 大暴雨   雨大 的 连 眼睛 都 睁不开   刘青 继续 说 着   瞬间 他们 就 被 雨 透 了   但 女子 依然 不 愿意 接受 刘青 的 帮助   就 继续 冒 着 大雨 往前走   大概 走 了 有 四十分钟 吧   女子 突然 停下 来说   我 想 回家 了   然后 女子 也 接受 了 刘青 递过来 的 小 黄衣   就 出现 了 深圳 微博上 的 照片   女子 披着 小 黄衣   刘青 小心翼翼 地 在 旁边 走 着 的 场景   从 南平 快速 下来 后   刘青 和 巡防 员 将 女子 带到 了 附近 的 坂田 派出所   那 姑娘 到底 是 遭遇 了 什么样 的 事情   才 会 说   要是 你 也 遭遇 我 的 事   你 也 会 这样   据 警方 透露   该 女子 姓 陈   系 湖北 人   今年 44 岁   据 家属 反映 其有 精神病 史   三天 前   陈某 从 老家 来 深圳 约会 网友   但 约会 受挫 导致 情绪 异常   女子 遂 产生 轻生 念头   目前   陈某 已经 被 送往 深圳 某 精神病 医院 进行 治疗   大大 君 只 希望 姑娘 能 早点 康复   其实 真爱 的 到来 并 不 存在 年龄 的 限制   你们 说 呢   因 善良 的 原因   一众 网友 纷纷 为   交警 暖 男点 ZAN   弓常 yan 桦   就 想 问 这个 小 哥哥 有 女票 吗   原谅 我 这 一辈子 浪荡 不羁 爱 萨摩耶   有 什么 过不去 的 要 轻生 嘛   想想 自己 的 家人   同时 也 感谢 交警 蜀黍   火心 聆听 心灵   点赞 交警   中華云盾   警察   警察 就是 群众 最 需时 申出 援手   Tomchlee   蜀黍 帅   SJ   李赫海 i   这个 交警 很 暖 有 木有   男子 迷奸 网友 拍 418 个 视频   女方 从 20 岁 到 50 岁 不 等   去年 6 月 7 号 上午   淮安市 涟水县 公安局 刑警大队 突然 接到 了 一个 奇怪 的 报警 电话   一名 女子 言语 不清   声称 自己 遭到 了 侵害   女子   被 侵害   言语 不清   几个 关键词 令 接到 电话 的 民警 瞬间 紧张 起来'"
      ]
     },
     "execution_count": 120,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_content[4]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Help on method generate_from_frequencies in module wordcloud.wordcloud:\n",
      "\n",
      "generate_from_frequencies(frequencies, max_font_size=None) method of wordcloud.wordcloud.WordCloud instance\n",
      "    Create a word_cloud from words and frequencies.\n",
      "    \n",
      "    Parameters\n",
      "    ----------\n",
      "    frequencies : dict from string to float\n",
      "        A contains words and associated frequency.\n",
      "    \n",
      "    max_font_size : int\n",
      "        Use this font-size instead of self.max_font_size\n",
      "    \n",
      "    Returns\n",
      "    -------\n",
      "    self\n",
      "\n"
     ]
    }
   ],
   "source": [
    "help(wc.generate_from_frequencies)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [],
   "source": [
    "machine_new_keywords_dict = {w: score for w, score in machine_new_keywords}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x15631149dd8>"
      ]
     },
     "execution_count": 125,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(wc.generate_from_frequencies(machine_new_keywords_dict))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [],
   "source": [
    "shenzhen_social_news = get_keywords_of_a_ducment(news_content[4])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('刘青', 67.12860103609216),\n",
       " ('女子', 49.805410612821845),\n",
       " ('南坪', 23.465811831805492),\n",
       " ('交警', 15.805787932978273),\n",
       " ('坂田', 13.953993984385477),\n",
       " ('黄衣', 13.050903997393533),\n",
       " ('巡防', 11.328810194210376),\n",
       " ('了', 11.012005636840637),\n",
       " ('轻生', 10.382998889874603),\n",
       " ('护栏', 9.32178985133465),\n",
       " ('坂雪岗', 9.302662656256985),\n",
       " ('裸身', 9.302662656256985),\n",
       " ('她', 9.285022817533253),\n",
       " ('约会', 9.267115391016052),\n",
       " ('深圳', 9.139285494790956),\n",
       " ('发现', 8.685059498920872),\n",
       " ('蜀黍', 8.214526567556433),\n",
       " ('网友', 8.126865289542744),\n",
       " ('突然', 8.047379355439626),\n",
       " ('十多分钟', 7.746360155489698),\n",
       " ('内裤', 7.746360155489698),\n",
       " ('着', 7.69376356788892),\n",
       " ('铁骑', 7.496482682273098),\n",
       " ('大概', 7.478906508099729),\n",
       " ('缓慢', 7.465405004419699),\n",
       " ('龙岗', 7.181266975549762),\n",
       " ('快速路', 7.108842630240872),\n",
       " ('大道', 6.863158044709045),\n",
       " ('环城', 6.816586558884396),\n",
       " ('辅警', 6.7921176460503725),\n",
       " ('往前走', 6.722593433531949),\n",
       " ('走', 6.658635364608071),\n",
       " ('视频', 6.65823837544899),\n",
       " ('市民', 6.50820380832565),\n",
       " ('精神病', 6.4895822953890745),\n",
       " ('主干道', 6.456170908383369),\n",
       " ('快速', 6.427950372933541),\n",
       " ('南路', 6.319939268588439),\n",
       " ('陈某', 6.319939268588439),\n",
       " ('情绪', 6.318072634988005),\n",
       " ('接到', 6.212930001065702),\n",
       " ('十字路口', 6.19005765472241),\n",
       " ('车流', 6.178076927344035),\n",
       " ('继续', 6.175118705343049),\n",
       " ('言语', 6.1662592081229946),\n",
       " ('劝说', 6.045884796156362),\n",
       " ('围观', 5.724912424706151),\n",
       " ('中队', 5.676835942971274),\n",
       " ('你', 5.665232674533076),\n",
       " ('停下', 5.6570193656507755),\n",
       " ('要是', 5.4856926184996855),\n",
       " ('还原', 5.475034951489551),\n",
       " ('遭遇', 5.428977234164574),\n",
       " ('这样', 5.38177119137673),\n",
       " ('观望', 5.281214925473439),\n",
       " ('不清', 5.170010805532909),\n",
       " ('接受', 5.151253446183707),\n",
       " ('侵害', 5.122852433378189),\n",
       " ('迷奸', 4.9523613237924735),\n",
       " ('镜头', 4.899868407616082),\n",
       " ('方向', 4.889185836300751),\n",
       " ('衣服', 4.826570449999394),\n",
       " ('也', 4.750174557013863),\n",
       " ('岁', 4.744091394164745),\n",
       " ('早上', 4.7270591963965325),\n",
       " ('此时', 4.696270541416007),\n",
       " ('姑娘', 4.662370084034877),\n",
       " ('火心', 4.651331328128492),\n",
       " ('刘青停', 4.651331328128492),\n",
       " ('地逆', 4.651331328128492),\n",
       " ('女票', 4.651331328128492),\n",
       " ('劝止', 4.651331328128492),\n",
       " ('男点', 4.651331328128492),\n",
       " ('ZAN', 4.651331328128492),\n",
       " ('涟水县', 4.651331328128492),\n",
       " ('刘青递', 4.651331328128492),\n",
       " ('浪荡', 4.651331328128492),\n",
       " ('披着', 4.651331328128492),\n",
       " ('中華云盾', 4.651331328128492),\n",
       " ('Tomchlee', 4.651331328128492),\n",
       " ('李赫海', 4.651331328128492),\n",
       " ('萨摩耶', 4.651331328128492),\n",
       " ('一辅警', 4.651331328128492),\n",
       " ('申出', 4.651331328128492),\n",
       " ('弓常', 4.651331328128492),\n",
       " ('SJ', 4.475240069072811),\n",
       " ('需时', 4.475240069072811),\n",
       " ('瞬间', 4.449640133527835),\n",
       " ('时', 4.428765345685334),\n",
       " ('一边', 4.413012257237489),\n",
       " ('下起', 4.350301332464511),\n",
       " ('往下跳', 4.350301332464511),\n",
       " ('四十分钟', 4.350301332464511),\n",
       " ('一丝不挂', 4.350301332464511),\n",
       " ('到底', 4.269591908465386),\n",
       " ('已经', 4.260101573573226),\n",
       " ('木有', 4.253391319456455),\n",
       " ('辅道', 4.253391319456455),\n",
       " ('睁不开', 4.253391319456455),\n",
       " ('并', 4.111278428408116),\n",
       " ('手要', 4.107263283778217),\n",
       " ('但', 4.098907380663682),\n",
       " ('险象环生', 4.04927133680053),\n",
       " ('大清早', 4.04927133680053),\n",
       " ('地往', 4.04927133680053),\n",
       " ('跳下去', 4.04927133680053),\n",
       " ('被', 4.033649542510729),\n",
       " ('路上', 4.002047609993112),\n",
       " ('拉住', 3.9981188143531488),\n",
       " ('排水沟', 3.9523613237924735),\n",
       " ('雨大', 3.9523613237924735),\n",
       " ('没', 3.9396278224389314),\n",
       " ('很', 3.931408900783627),\n",
       " ('我', 3.920203291368041),\n",
       " ('不羁', 3.9109686386342486),\n",
       " ('真爱', 3.9109686386342486),\n",
       " ('递过来', 3.9109686386342486),\n",
       " ('的', 3.890424627403554),\n",
       " ('微博称', 3.873180077744849),\n",
       " ('44', 3.8706559689873865),\n",
       " ('呢', 3.8573948111893603),\n",
       " ('其有', 3.838417971485637),\n",
       " ('刑警大队', 3.838417971485637),\n",
       " ('愿意', 3.804736933744662),\n",
       " ('警察', 3.800874815492734),\n",
       " ('值勤', 3.776270064736792),\n",
       " ('女方', 3.776270064736792),\n",
       " ('出现', 3.7560113879094406),\n",
       " ('过不去', 3.748241341136549),\n",
       " ('开导', 3.7219124024142),\n",
       " ('穿衣服', 3.7219124024142),\n",
       " ('yan', 3.7219124024142),\n",
       " ('沿路', 3.6970888186891675),\n",
       " ('别扭', 3.6970888186891675),\n",
       " ('赶忙', 3.6970888186891675),\n",
       " ('就', 3.6878522085808467),\n",
       " ('走时', 3.6736077228396447),\n",
       " ('纷纷', 3.6382837341199585),\n",
       " ('冲上去', 3.6301420290585544),\n",
       " ('办法', 3.611846377013398),\n",
       " ('逆行', 3.6099386429702673),\n",
       " ('挣脱', 3.5721500820808676),\n",
       " ('略带', 3.5721500820808676),\n",
       " ('暖', 3.567727680538882),\n",
       " ('碍于', 3.554421315120436),\n",
       " ('原谅', 3.5373879758216558),\n",
       " ('不', 3.5242870574887766),\n",
       " ('赤裸', 3.4752400690728114),\n",
       " ('尾随', 3.4752400690728114),\n",
       " ('警情', 3.4752400690728114),\n",
       " ('帅气', 3.4472113454725677),\n",
       " ('依然', 3.441273880927914),\n",
       " ('排起', 3.4208824067502186),\n",
       " ('418', 3.4208824067502186),\n",
       " ('低落', 3.408293279442198),\n",
       " ('警方', 3.4082855252047977),\n",
       " ('都', 3.4079467568281907),\n",
       " ('电话', 3.3864909592148145),\n",
       " ('驱赶', 3.3503013324645115),\n",
       " ('一众', 3.3503013324645115),\n",
       " ('放慢', 3.3503013324645115),\n",
       " ('小', 3.3501691227697368),\n",
       " ('早点', 3.339577467072738),\n",
       " ('南平', 3.339577467072738),\n",
       " ('援手', 3.329112033394573),\n",
       " ('淡定', 3.29914881001713),\n",
       " ('上身', 3.2711200864168863),\n",
       " ('腼腆', 3.2533913194564548),\n",
       " ('再次出现', 3.2363579801576745),\n",
       " ('意想不到', 3.2280854541916844),\n",
       " ('桦', 3.2041732967862733),\n",
       " ('受挫', 3.196486468119982),\n",
       " ('淮安市', 3.188933330229536),\n",
       " ('从', 3.1671096611590173),\n",
       " ('披上', 3.1461813498085864),\n",
       " ('一位', 3.106760514268685),\n",
       " ('奔走', 3.095028827361205),\n",
       " ('挡住', 3.0890384636720176),\n",
       " ('善良', 3.060266721101993),\n",
       " ('想想', 3.054734232502032),\n",
       " ('小心翼翼', 3.04927133680053),\n",
       " ('长队', 3.028082037730592),\n",
       " ('大暴雨', 3.0128420711738553),\n",
       " ('微博上', 3.0128420711738553),\n",
       " ('裸露', 3.0029713171475607),\n",
       " ('外表', 3.0029713171475607),\n",
       " ('附近', 3.0014567526709657),\n",
       " ('一路上', 2.9885734964469184),\n",
       " ('冲向', 2.9885734964469184),\n",
       " ('念头', 2.9838783752385383),\n",
       " ('还', 2.946049390724077),\n",
       " ('骑着', 2.9353279844936933),\n",
       " ('群众', 2.9332797946217872),\n",
       " ('嘛', 2.9270554585277035),\n",
       " ('什么', 2.9264939088271085),\n",
       " ('奇怪', 2.899282880309054),\n",
       " ('上前', 2.8695759534760237),\n",
       " ('年头', 2.812482237391237),\n",
       " ('安慰', 2.812482237391237),\n",
       " ('而行', 2.7880084680080364),\n",
       " ('女生', 2.759236725438012),\n",
       " ('高峰期', 2.75646167138324),\n",
       " ('高大', 2.72964485264489),\n",
       " ('赶紧', 2.714315220663678),\n",
       " ('聆听', 2.704388057430667),\n",
       " ('期间', 2.687029002932124),\n",
       " ('驻足', 2.678203474528794),\n",
       " ('全身', 2.6736077228396447),\n",
       " ('有', 2.6584688103348446),\n",
       " ('一辈子', 2.649165266371985),\n",
       " ('哥哥', 2.6470099543458496),\n",
       " ('机动车', 2.6219475504432825),\n",
       " ('老家', 2.619922863876868),\n",
       " ('心灵', 2.617907572641543),\n",
       " ('朝着', 2.6099386429702673),\n",
       " ('现场', 2.605858201101714),\n",
       " ('路边', 2.6001788056811113),\n",
       " ('带到', 2.573963422844336),\n",
       " ('举动', 2.5685459578120424),\n",
       " ('扔', 2.5667550501941614),\n",
       " ('恐怕', 2.5631952394279414),\n",
       " ('行进', 2.557909642966257),\n",
       " ('尴尬', 2.554421315120436),\n",
       " ('大雨', 2.5390615597112216),\n",
       " ('遂', 2.5258500624278986),\n",
       " ('穿着', 2.517792419758275),\n",
       " ('旁边', 2.509881554728025),\n",
       " ('远离', 2.499042983745436),\n",
       " ('躺', 2.4975164637839633),\n",
       " ('声称', 2.4975164637839633),\n",
       " ('货车', 2.492968836033243),\n",
       " ('说话', 2.486978472344055),\n",
       " ('点赞', 2.485493703438364),\n",
       " ('派出所', 2.472354380835323),\n",
       " ('脚步', 2.460999629958201),\n",
       " ('这个', 2.4579755135469785),\n",
       " ('靠近', 2.454050770002873),\n",
       " ('后面', 2.4526742411740696),\n",
       " ('盯', 2.449934203808041),\n",
       " ('沿着', 2.4351654258424995),\n",
       " ('上面', 2.433847383914586),\n",
       " ('什么样', 2.4312232400884373),\n",
       " ('耐心', 2.4208824067502186),\n",
       " ('紧紧', 2.409535896833294),\n",
       " ('白色', 2.409535896833294),\n",
       " ('上班', 2.409535896833294),\n",
       " ('重复', 2.3984782971485994),\n",
       " ('暴雨', 2.3912599401434176),\n",
       " ('上去', 2.387695259540384),\n",
       " ('不时', 2.3680300994249426),\n",
       " ('跟随', 2.3680300994249426),\n",
       " ('送往', 2.362411722466766),\n",
       " ('报警', 2.3612967167659744),\n",
       " ('眼睛', 2.3590752567720163),\n",
       " ('异常', 2.3590752567720163),\n",
       " ('您', 2.3568651019668994),\n",
       " ('正值', 2.3481352707080037),\n",
       " ('清晨', 2.3406380157851316),\n",
       " ('康复', 2.328079227956805),\n",
       " ('事发', 2.307922734324635),\n",
       " ('不敢', 2.298184781914513),\n",
       " ('刺激', 2.2849083709025195),\n",
       " ('一名', 2.2786287172647315),\n",
       " ('情况', 2.273036066258414),\n",
       " ('见到', 2.2648323625778395),\n",
       " ('你们', 2.2630524646688532),\n",
       " ('家属', 2.2603962210251134),\n",
       " ('行走', 2.2595144045152438),\n",
       " ('三天', 2.258634374868827),\n",
       " ('一句', 2.2499307873469485),\n",
       " ('听到', 2.226449691497425),\n",
       " ('往', 2.2235068895663668),\n",
       " ('心里', 2.2231965340997037),\n",
       " ('昨日', 2.2191640586859043),\n",
       " ('自己', 2.214774771291172),\n",
       " ('公安局', 2.198778264899567),\n",
       " ('民警', 2.196486468119982),\n",
       " ('周围', 2.1904334853719445),\n",
       " ('场景', 2.1896827600650375),\n",
       " ('话', 2.179267590948998),\n",
       " ('想', 2.1774822070443163),\n",
       " ('回家', 2.1627806116280484),\n",
       " ('到', 2.1497523584585987),\n",
       " ('监控', 2.1455032942736563),\n",
       " ('大队', 2.133475909198464),\n",
       " ('聚集', 2.1308477953877003),\n",
       " ('没有', 2.1292630530088865),\n",
       " ('上', 2.1219500466542325),\n",
       " ('试图', 2.1004917230627074),\n",
       " ('感谢', 2.095028827361205),\n",
       " ('开始', 2.092052563974766),\n",
       " ('遭到', 2.0825431158131456),\n",
       " ('冒', 2.0744143721632855),\n",
       " ('中间', 2.073839528291267),\n",
       " ('大大', 2.0613817268027845),\n",
       " ('说', 2.049245802061725),\n",
       " ('挺', 2.0369614885802036),\n",
       " ('君', 2.034855814239927),\n",
       " ('身上', 2.0244779134617668),\n",
       " ('尝试', 2.0193805018692754),\n",
       " ('危险', 1.9899919881224526),\n",
       " ('吧', 1.9545382430467482),\n",
       " ('然后', 1.945895281543242),\n",
       " ('年龄', 1.943761152030556),\n",
       " ('一辆', 1.9420613671526616),\n",
       " ('治疗', 1.9258280592691768),\n",
       " ('拒', 1.920145620494492),\n",
       " ('家人', 1.9137411618427709),\n",
       " ('牵', 1.9101797292767073),\n",
       " ('寻找', 1.9089990457713442),\n",
       " ('只', 1.908029572792218),\n",
       " ('一只', 1.9046971291909136),\n",
       " ('事情', 1.9027557111975004),\n",
       " ('紧张', 1.8988987188670183),\n",
       " ('离开', 1.8834337121104017),\n",
       " ('到来', 1.8823234571847185),\n",
       " ('吗', 1.8728183163895675),\n",
       " ('感觉', 1.8702943895073605),\n",
       " ('车辆', 1.8649348667561882),\n",
       " ('限制', 1.864579905982931),\n",
       " ('根本', 1.8575409434376737),\n",
       " ('反映', 1.833435570966697),\n",
       " ('几个', 1.831129868642852),\n",
       " ('该', 1.8266086115180598),\n",
       " ('透露', 1.8255807467804646),\n",
       " ('地', 1.8182999882820219),\n",
       " ('关键词', 1.8165928092246513),\n",
       " ('帅', 1.8003789283349996),\n",
       " ('其实', 1.7994617273987261),\n",
       " ('速度', 1.7868202470701005),\n",
       " ('只能', 1.7820996083975162),\n",
       " ('些', 1.78159152381527),\n",
       " ('那么', 1.7620296256221821),\n",
       " ('去', 1.7578725340929844),\n",
       " ('出口', 1.747969991873174),\n",
       " ('姓', 1.7401737193885158),\n",
       " ('一份', 1.739907362752198),\n",
       " ('下来', 1.7186039608269632),\n",
       " ('另外', 1.6973301604469222),\n",
       " ('湖北', 1.6731508111910784),\n",
       " ('上午', 1.669965819049948),\n",
       " ('青年', 1.659662320748544),\n",
       " ('随后', 1.645721882768212),\n",
       " ('或者', 1.632422883812165),\n",
       " ('医院', 1.6286967881843735),\n",
       " ('面对', 1.6211365427717412),\n",
       " ('当时', 1.6157015003380535),\n",
       " ('大家', 1.603861753508636),\n",
       " ('作出', 1.6034726007210356),\n",
       " ('某', 1.5996001320686428),\n",
       " ('36', 1.5992147775784944),\n",
       " ('后', 1.5830759646797947),\n",
       " ('90', 1.5796333829068783),\n",
       " ('来说', 1.5787138515832557),\n",
       " ('产生', 1.5759668817552075),\n",
       " ('起来', 1.5545950676660236),\n",
       " ('控制', 1.5525146110795511),\n",
       " ('坐', 1.5492408026166558),\n",
       " ('穿', 1.5330666020390131),\n",
       " ('和', 1.5243340291611516),\n",
       " ('但是', 1.5230936573593052),\n",
       " ('标题', 1.5214803773395824),\n",
       " ('i', 1.5036550038873937),\n",
       " ('南', 1.499637037997105),\n",
       " ('导致', 1.4940256978005262),\n",
       " ('跑', 1.4923660677450823),\n",
       " ('原因', 1.4901131084374761),\n",
       " ('跟', 1.481069612733535),\n",
       " ('看到', 1.4469399962091927),\n",
       " ('不少', 1.4308309825137626),\n",
       " ('关于', 1.423958885838856),\n",
       " ('雨', 1.4122822349883009),\n",
       " ('为了', 1.4080451820450464),\n",
       " ('会', 1.399238623728092),\n",
       " ('16', 1.3983375078280595),\n",
       " ('存在', 1.3926940454044159),\n",
       " ('追', 1.3887615949109426),\n",
       " ('事', 1.3790139567525608),\n",
       " ('帮助', 1.3738367185174805),\n",
       " ('男子', 1.3700706410734795),\n",
       " ('员', 1.3697851181993932),\n",
       " ('一起', 1.3670134126978826),\n",
       " ('交流', 1.3570856119903743),\n",
       " ('据', 1.3432473005755838),\n",
       " ('这一', 1.332850602953975),\n",
       " ('令', 1.3271517984105927),\n",
       " ('其', 1.3209198466540129),\n",
       " ('停', 1.3098968035503522),\n",
       " ('去年', 1.2942544882860807),\n",
       " ('就是', 1.2374450757988886),\n",
       " ('不断', 1.235690348232338),\n",
       " ('一路', 1.2244948743249844),\n",
       " ('希望', 1.2148506331189974),\n",
       " ('住', 1.197020139980926),\n",
       " ('正在', 1.1947375778840845),\n",
       " ('手', 1.1873304484373264),\n",
       " ('发生', 1.1824672879803046),\n",
       " ('了解', 1.1690295609050496),\n",
       " ('50', 1.1678866800299572),\n",
       " ('为', 1.1607802604541433),\n",
       " ('要', 1.1560692992858268),\n",
       " ('爱', 1.1554402484623048),\n",
       " ('可能', 1.1455032942736563),\n",
       " ('陈', 1.138313602796874),\n",
       " ('当', 1.1360880308917092),\n",
       " ('那', 1.1350604453991524),\n",
       " ('史', 1.1277149090741216),\n",
       " ('号', 1.1255740849761815),\n",
       " ('拍', 1.1204915495119718),\n",
       " ('25', 1.1140160550164828),\n",
       " ('才', 1.1139530452982869),\n",
       " ('进入', 1.1128832764182752),\n",
       " ('介绍', 1.1120669921810373),\n",
       " ('他们', 1.1044803263470984),\n",
       " ('可以', 1.0866652638764032),\n",
       " ('花', 1.0686998886388561),\n",
       " ('提供', 1.0659270041635354),\n",
       " ('把', 1.059432341437268),\n",
       " ('连', 1.0414169180424948),\n",
       " ('人', 1.0398762897471812),\n",
       " ('将', 1.0305857067286253),\n",
       " ('同时', 1.0135915010093561),\n",
       " ('分', 0.9973828174544938),\n",
       " ('透', 0.99720293686668),\n",
       " ('今年', 0.966216859081953),\n",
       " ('记者', 0.9271998038792411),\n",
       " ('让', 0.9137014294899777),\n",
       " ('目前', 0.8933662302670573),\n",
       " ('问', 0.879670368779605),\n",
       " ('看', 0.8770699049091711),\n",
       " ('原', 0.8745938334104167),\n",
       " ('表示', 0.8634105895744374),\n",
       " ('一个', 0.8556598221824707),\n",
       " ('因', 0.8181204478458837),\n",
       " ('系', 0.7946628445169575),\n",
       " ('带', 0.7763279745284514),\n",
       " ('6', 0.7622602583437212),\n",
       " ('好', 0.7288653828300791),\n",
       " ('拉', 0.7275952497445629),\n",
       " ('进行', 0.7263804392128818),\n",
       " ('起', 0.7202295824958537),\n",
       " ('一', 0.710829003592252),\n",
       " ('他', 0.7074907017272314),\n",
       " ('正', 0.6973060202350319),\n",
       " ('可', 0.6842515939839953),\n",
       " ('最', 0.6400993178091264),\n",
       " ('天', 0.6235944942009486),\n",
       " ('能', 0.6185111786899286),\n",
       " ('里', 0.6039006879729503),\n",
       " ('等', 0.5441043941837564),\n",
       " ('中', 0.5156135775503206),\n",
       " ('个', 0.5035468301229501),\n",
       " ('这', 0.5000074536188772),\n",
       " ('与', 0.49074275983845767),\n",
       " ('多', 0.4809962583744546),\n",
       " ('前', 0.4735784562858743),\n",
       " ('7', 0.45891756623126934),\n",
       " ('来', 0.45742792557574563),\n",
       " ('是', 0.4069428640413574),\n",
       " ('在', 0.32985711253084343),\n",
       " ('大', 0.3132734527087363),\n",
       " ('照片', 0.24212545393891896),\n",
       " ('20', 0.1475474522482),\n",
       " ('月', 0.06974313771994531),\n",
       " ('日', 0.05085505597918966)]"
      ]
     },
     "execution_count": 127,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "shenzhen_social_news"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [],
   "source": [
    "from PIL import Image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [],
   "source": [
    "police_mask = np.array(Image.open('./data/mask.png'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1 1 1 ... 1 1 1]\n",
      " [1 1 1 ... 1 1 1]\n",
      " [1 1 1 ... 1 1 1]\n",
      " ...\n",
      " [1 1 1 ... 1 1 1]\n",
      " [1 1 1 ... 1 1 1]\n",
      " [1 1 1 ... 1 1 1]]\n"
     ]
    }
   ],
   "source": [
    "print(police_mask)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 135,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "os.path.exists(\"./data/mask.png\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [],
   "source": [
    "wordcloud_with_mask = wordcloud.WordCloud(\n",
    "font_path='/Users/mqgao/Downloads/SourceHanSerifSC-Regular.otf', \n",
    "mask=police_mask\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x156312e2e10>"
      ]
     },
     "execution_count": 133,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(wc.generate_from_frequencies({w:s for w, s in shenzhen_social_news[:20]}))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {},
   "outputs": [
    {
     "ename": "OSError",
     "evalue": "cannot open resource",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mOSError\u001b[0m                                   Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-138-a1866ba34bbc>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0mplt\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mimshow\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mwordcloud_with_mask\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mgenerate_from_frequencies\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m{\u001b[0m\u001b[0mw\u001b[0m\u001b[1;33m:\u001b[0m\u001b[0ms\u001b[0m \u001b[1;32mfor\u001b[0m \u001b[0mw\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0ms\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mshenzhen_social_news\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;36m20\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m}\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[1;32m~\\AppData\\Local\\Continuum\\anaconda3\\envs\\cs231n\\lib\\site-packages\\wordcloud\\wordcloud.py\u001b[0m in \u001b[0;36mgenerate_from_frequencies\u001b[1;34m(self, frequencies, max_font_size)\u001b[0m\n\u001b[0;32m    424\u001b[0m             \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    425\u001b[0m                 self.generate_from_frequencies(dict(frequencies[:2]),\n\u001b[1;32m--> 426\u001b[1;33m                                                max_font_size=self.height)\n\u001b[0m\u001b[0;32m    427\u001b[0m                 \u001b[1;31m# find font sizes\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    428\u001b[0m                 \u001b[0msizes\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;33m[\u001b[0m\u001b[0mx\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m]\u001b[0m \u001b[1;32mfor\u001b[0m \u001b[0mx\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mlayout_\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Continuum\\anaconda3\\envs\\cs231n\\lib\\site-packages\\wordcloud\\wordcloud.py\u001b[0m in \u001b[0;36mgenerate_from_frequencies\u001b[1;34m(self, frequencies, max_font_size)\u001b[0m\n\u001b[0;32m    471\u001b[0m             \u001b[1;32mwhile\u001b[0m \u001b[1;32mTrue\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    472\u001b[0m                 \u001b[1;31m# try to find a position\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 473\u001b[1;33m                 \u001b[0mfont\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mImageFont\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtruetype\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfont_path\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mfont_size\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    474\u001b[0m                 \u001b[1;31m# transpose font optionally\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    475\u001b[0m                 transposed_font = ImageFont.TransposedFont(\n",
      "\u001b[1;32m~\\AppData\\Local\\Continuum\\anaconda3\\envs\\cs231n\\lib\\site-packages\\PIL\\ImageFont.py\u001b[0m in \u001b[0;36mtruetype\u001b[1;34m(font, size, index, encoding, layout_engine)\u001b[0m\n\u001b[0;32m    259\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    260\u001b[0m     \u001b[1;32mtry\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 261\u001b[1;33m         \u001b[1;32mreturn\u001b[0m \u001b[0mFreeTypeFont\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfont\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0msize\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mindex\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mencoding\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlayout_engine\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    262\u001b[0m     \u001b[1;32mexcept\u001b[0m \u001b[0mIOError\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    263\u001b[0m         \u001b[0mttf_filename\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mos\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpath\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mbasename\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfont\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Continuum\\anaconda3\\envs\\cs231n\\lib\\site-packages\\PIL\\ImageFont.py\u001b[0m in \u001b[0;36m__init__\u001b[1;34m(self, font, size, index, encoding, layout_engine)\u001b[0m\n\u001b[0;32m    142\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    143\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0misPath\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfont\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 144\u001b[1;33m             \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfont\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mcore\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mgetfont\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfont\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0msize\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mindex\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mencoding\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlayout_engine\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mlayout_engine\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    145\u001b[0m         \u001b[1;32melse\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    146\u001b[0m             \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfont_bytes\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mfont\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mread\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mOSError\u001b[0m: cannot open resource"
     ]
    }
   ],
   "source": [
    "plt.imshow(wordcloud_with_mask.generate_from_frequencies({w:s for w, s in shenzhen_social_news[:20]}))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## TFIDF Vectorizezd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.feature_extraction.text import TfidfVectorizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [],
   "source": [
    "vectorized = TfidfVectorizer(max_features=10000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['此外   自 本周   6 月 12 日   起   除 小米 手机 6 等 15 款 机型 外   其余 机型 已 暂停 更新 发布   含 开发 版   体验版 内测   稳定版 暂不受 影响   以 确保 工程师 可以 集中 全部 精力 进行 系统优化 工作   有人 猜测 这 也 是 将 精力 主要 用到 MIUI   9 的 研发 之中   MIUI   8 去年 5 月 发布   距今已有 一年 有余   也 是 时候 更新换代 了   当然   关于 MIUI   9 的 确切 信息   我们 还是 等待 官方消息']"
      ]
     },
     "execution_count": 146,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_content[:1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {},
   "outputs": [],
   "source": [
    "sample_num = 50000\n",
    "sub_samples = news_content[:sample_num]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {},
   "outputs": [],
   "source": [
    "X = vectorized.fit_transform(sub_samples)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(50000, 10000)"
      ]
     },
     "execution_count": 147,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'此外': 6195,\n",
       " '本周': 5892,\n",
       " '12': 35,\n",
       " '小米': 3974,\n",
       " '手机': 4878,\n",
       " '15': 46,\n",
       " '其余': 1844,\n",
       " '暂停': 5738,\n",
       " '更新': 5760,\n",
       " '发布': 2635,\n",
       " '开发': 4405,\n",
       " '影响': 4525,\n",
       " '确保': 7338,\n",
       " '工程师': 4132,\n",
       " '可以': 2734,\n",
       " '集中': 9631,\n",
       " '全部': 1743,\n",
       " '精力': 7662,\n",
       " '进行': 9141,\n",
       " '工作': 4113,\n",
       " '有人': 5808,\n",
       " '猜测': 6880,\n",
       " '主要': 912,\n",
       " '研发': 7319,\n",
       " '之中': 933,\n",
       " '去年': 2558,\n",
       " '一年': 302,\n",
       " '时候': 5628,\n",
       " '当然': 4505,\n",
       " '关于': 1819,\n",
       " '信息': 1546,\n",
       " '我们': 4802,\n",
       " '还是': 9082,\n",
       " '等待': 7607,\n",
       " '作为': 1440,\n",
       " '唯一': 2966,\n",
       " '通过': 9249,\n",
       " '10': 12,\n",
       " '平台': 4255,\n",
       " '认证': 8554,\n",
       " '处理器': 3331,\n",
       " '强调': 4481,\n",
       " '不会': 551,\n",
       " '因为': 3045,\n",
       " '考虑': 7980,\n",
       " '性能': 4642,\n",
       " '核心': 6069,\n",
       " '相反': 7212,\n",
       " '他们': 1242,\n",
       " '联手': 8020,\n",
       " '微软': 4579,\n",
       " '找到': 4949,\n",
       " '一种': 344,\n",
       " '适合': 9193,\n",
       " '兼顾': 1864,\n",
       " '完美': 3783,\n",
       " '方案': 5537,\n",
       " '报道': 5026,\n",
       " '已经': 4156,\n",
       " '拿到': 5092,\n",
       " '一些': 253,\n",
       " '以便': 1259,\n",
       " '更好': 5757,\n",
       " '理解': 6966,\n",
       " '架构': 6022,\n",
       " '资料': 8791,\n",
       " '显示': 5703,\n",
       " '一款': 332,\n",
       " '集成': 9638,\n",
       " '传统': 1355,\n",
       " '节省': 8201,\n",
       " '至少': 8132,\n",
       " '30': 103,\n",
       " '空间': 7498,\n",
       " '按计划': 5124,\n",
       " '今年': 1217,\n",
       " '联想': 8019,\n",
       " '首发': 9825,\n",
       " '电脑': 7067,\n",
       " '预计': 9745,\n",
       " '形态': 4513,\n",
       " '产品': 1135,\n",
       " '只是': 2728,\n",
       " '开始': 4412,\n",
       " '未来': 5886,\n",
       " '也许': 982,\n",
       " '见到': 8475,\n",
       " '三星': 443,\n",
       " '华为': 2417,\n",
       " '澎湃': 6699,\n",
       " '进入': 9123,\n",
       " '此前': 6193,\n",
       " '搭载': 5252,\n",
       " '电池': 7063,\n",
       " '规格': 8495,\n",
       " '至于': 8129,\n",
       " '可能': 2741,\n",
       " '所说': 4870,\n",
       " '设定': 8585,\n",
       " '有关': 5810,\n",
       " '按照': 5122,\n",
       " '目前': 7188,\n",
       " '掌握': 5160,\n",
       " '拥有': 5083,\n",
       " '摄像头': 5256,\n",
       " '惊喜': 4717,\n",
       " '根据': 6078,\n",
       " '京东': 1144,\n",
       " '泄露': 6406,\n",
       " '售价': 2964,\n",
       " '应该': 4344,\n",
       " '某个': 6025,\n",
       " '这是': 9105,\n",
       " '18': 53,\n",
       " '葡萄牙': 8322,\n",
       " '中部': 858,\n",
       " '地区': 3168,\n",
       " '拍摄': 5064,\n",
       " '森林': 6119,\n",
       " '大火': 3485,\n",
       " '汽车': 6355,\n",
       " '新华社': 5489,\n",
       " '记者': 8569,\n",
       " '标题': 6055,\n",
       " '44': 125,\n",
       " '女子': 3625,\n",
       " '深圳': 6576,\n",
       " '网友': 7888,\n",
       " '暴雨': 5748,\n",
       " '交警': 1120,\n",
       " '昨日': 5694,\n",
       " '清晨': 6604,\n",
       " '发现': 2647,\n",
       " '行走': 8392,\n",
       " '快速': 4622,\n",
       " '期间': 5881,\n",
       " '赶紧': 8852,\n",
       " '一路': 365,\n",
       " '那么': 9299,\n",
       " '事发': 1024,\n",
       " '到底': 2154,\n",
       " '发生': 2649,\n",
       " '什么': 1208,\n",
       " '一起': 364,\n",
       " '还原': 9081,\n",
       " '现场': 6927,\n",
       " '大队': 3517,\n",
       " '中队': 863,\n",
       " '女生': 3631,\n",
       " '一位': 259,\n",
       " '高大': 9919,\n",
       " '说话': 8668,\n",
       " '90': 184,\n",
       " '青年': 9672,\n",
       " '介绍': 1224,\n",
       " '16': 49,\n",
       " '早上': 5617,\n",
       " '36': 113,\n",
       " '正在': 6180,\n",
       " '附近': 9550,\n",
       " '接到': 5177,\n",
       " '机动车': 5928,\n",
       " '危险': 2511,\n",
       " '随后': 9602,\n",
       " '寻找': 3932,\n",
       " '大概': 3476,\n",
       " '大道': 3511,\n",
       " '出口': 1998,\n",
       " '方向': 5535,\n",
       " '身上': 8929,\n",
       " '另外': 2722,\n",
       " '一名': 276,\n",
       " '情绪': 4714,\n",
       " '尝试': 3996,\n",
       " '交流': 1115,\n",
       " '离开': 7403,\n",
       " '愿意': 4764,\n",
       " '接受': 5179,\n",
       " '继续': 7832,\n",
       " '缓慢': 7862,\n",
       " '此时': 6197,\n",
       " '路边': 8918,\n",
       " '聚集': 8034,\n",
       " '市民': 4192,\n",
       " '围观': 3065,\n",
       " '为了': 874,\n",
       " '刺激': 2179,\n",
       " '一边': 370,\n",
       " '群众': 7950,\n",
       " '警方': 8538,\n",
       " '提供': 5215,\n",
       " '一份': 257,\n",
       " '视频': 8509,\n",
       " '了解': 1007,\n",
       " '25': 95,\n",
       " '出现': 2021,\n",
       " '监控': 7165,\n",
       " '穿着': 7502,\n",
       " '白色': 7140,\n",
       " '沿着': 6403,\n",
       " '当时': 4502,\n",
       " '正值': 6179,\n",
       " '上班': 495,\n",
       " '高峰期': 9927,\n",
       " '路上': 8911,\n",
       " '纷纷': 7737,\n",
       " '驻足': 9885,\n",
       " '观望': 8485,\n",
       " '不少': 576,\n",
       " '车辆': 8955,\n",
       " '速度': 9253,\n",
       " '脚步': 8077,\n",
       " '依然': 1483,\n",
       " '行进': 8393,\n",
       " '中间': 861,\n",
       " '一辆': 368,\n",
       " '货车': 8755,\n",
       " '镜头': 9454,\n",
       " '但是': 1379,\n",
       " '全身': 1740,\n",
       " '朝着': 5865,\n",
       " '周围': 2907,\n",
       " '没有': 6373,\n",
       " '或者': 4811,\n",
       " '上前': 462,\n",
       " '看到': 7244,\n",
       " '这样': 9107,\n",
       " '情况': 4708,\n",
       " '恐怕': 4678,\n",
       " '大家': 3459,\n",
       " '办法': 2232,\n",
       " '面对': 9693,\n",
       " '这一': 9085,\n",
       " '表示': 8421,\n",
       " '根本': 6079,\n",
       " '不敢': 588,\n",
       " '心里': 4603,\n",
       " '感觉': 4760,\n",
       " '尴尬': 4014,\n",
       " '跟随': 8905,\n",
       " '作出': 1441,\n",
       " '举动': 919,\n",
       " '突然': 7511,\n",
       " '靠近': 9687,\n",
       " '上面': 510,\n",
       " '控制': 5190,\n",
       " '远离': 9146,\n",
       " '衣服': 8404,\n",
       " '只能': 2730,\n",
       " '紧紧': 7691,\n",
       " '一只': 272,\n",
       " '后面': 2867,\n",
       " '耐心': 7989,\n",
       " '安慰': 3766,\n",
       " '听到': 2879,\n",
       " '不断': 589,\n",
       " '重复': 9364,\n",
       " '一句': 271,\n",
       " '要是': 8468,\n",
       " '遭遇': 9288,\n",
       " '不时': 593,\n",
       " '试图': 8626,\n",
       " '大暴雨': 3472,\n",
       " '眼睛': 7270,\n",
       " '瞬间': 7288,\n",
       " '帮助': 4224,\n",
       " '大雨': 3519,\n",
       " '来说': 5995,\n",
       " '回家': 3029,\n",
       " '然后': 6774,\n",
       " '照片': 6781,\n",
       " '旁边': 5546,\n",
       " '场景': 3196,\n",
       " '下来': 525,\n",
       " '带到': 4212,\n",
       " '派出所': 6470,\n",
       " '姑娘': 3669,\n",
       " '什么样': 1209,\n",
       " '事情': 1029,\n",
       " '透露': 9219,\n",
       " '湖北': 6645,\n",
       " '家属': 3883,\n",
       " '反映': 2615,\n",
       " '三天': 437,\n",
       " '陈某': 9557,\n",
       " '老家': 7963,\n",
       " '导致': 3939,\n",
       " '异常': 4441,\n",
       " '产生': 1139,\n",
       " '送往': 9191,\n",
       " '医院': 2382,\n",
       " '治疗': 6400,\n",
       " '大大': 3452,\n",
       " '希望': 4207,\n",
       " '康复': 4363,\n",
       " '其实': 1846,\n",
       " '到来': 2156,\n",
       " '存在': 3707,\n",
       " '年龄': 4296,\n",
       " '限制': 9577,\n",
       " '你们': 1451,\n",
       " '原因': 2542,\n",
       " '这个': 9086,\n",
       " '哥哥': 2953,\n",
       " '一辈子': 369,\n",
       " '自己': 8099,\n",
       " '家人': 3878,\n",
       " '同时': 2831,\n",
       " '感谢': 4761,\n",
       " '聆听': 7996,\n",
       " '心灵': 4598,\n",
       " '点赞': 6745,\n",
       " '警察': 8534,\n",
       " '就是': 4005,\n",
       " '男子': 7080,\n",
       " '20': 64,\n",
       " '50': 132,\n",
       " '上午': 464,\n",
       " '公安局': 1771,\n",
       " '一个': 247,\n",
       " '报警': 5025,\n",
       " '电话': 7072,\n",
       " '声称': 3321,\n",
       " '遭到': 9286,\n",
       " '侵害': 1492,\n",
       " '几个': 1975,\n",
       " '关键词': 1834,\n",
       " '民警': 6287,\n",
       " '紧张': 7688,\n",
       " '起来': 8855,\n",
       " '受到': 2677,\n",
       " 'a股': 198,\n",
       " '纳入': 7730,\n",
       " 'msci': 219,\n",
       " '指数': 5115,\n",
       " '利好': 2136,\n",
       " '消息': 6544,\n",
       " '市场': 4178,\n",
       " '周三': 2901,\n",
       " '再度': 1884,\n",
       " '上演': 494,\n",
       " '行情': 8383,\n",
       " '周四': 2906,\n",
       " '金融股': 9409,\n",
       " '白马股': 7143,\n",
       " '表现': 8420,\n",
       " '尾盘': 4030,\n",
       " '跳水': 8922,\n",
       " '之后': 940,\n",
       " '金融': 9400,\n",
       " '板块': 6001,\n",
       " '状态': 6865,\n",
       " '分析': 2060,\n",
       " '人士': 1172,\n",
       " '认为': 8548,\n",
       " '受益': 2685,\n",
       " '估值': 1374,\n",
       " '资金': 8803,\n",
       " '青睐': 9679,\n",
       " '存量': 3711,\n",
       " '博弈': 2476,\n",
       " '格局': 6082,\n",
       " '风格': 9784,\n",
       " '震荡': 9662,\n",
       " '延续': 4373,\n",
       " '流动性': 6477,\n",
       " '改善': 5333,\n",
       " '经济': 7779,\n",
       " '预期': 9738,\n",
       " '有助于': 5814,\n",
       " '支撑': 5301,\n",
       " '大盘': 3490,\n",
       " '逐步': 9222,\n",
       " '再现': 1887,\n",
       " '未能': 5889,\n",
       " '上行': 501,\n",
       " '态势': 4629,\n",
       " '两市': 710,\n",
       " '成交': 4777,\n",
       " '小幅': 3964,\n",
       " '放量': 5367,\n",
       " '29': 102,\n",
       " '中信': 757,\n",
       " '一级': 352,\n",
       " '行业': 8375,\n",
       " '收盘': 5320,\n",
       " '银行': 9434,\n",
       " '两个': 688,\n",
       " '分别': 2045,\n",
       " '上涨': 490,\n",
       " '80': 171,\n",
       " '二级': 1048,\n",
       " '来看': 5991,\n",
       " '股份制': 8046,\n",
       " '涨幅': 6563,\n",
       " '最高': 5797,\n",
       " '达到': 9021,\n",
       " '22': 91,\n",
       " '82': 175,\n",
       " '信托': 1552,\n",
       " '及其': 2586,\n",
       " '64': 152,\n",
       " '保险': 1536,\n",
       " '34': 109,\n",
       " '证券': 8598,\n",
       " '06': 8,\n",
       " '成分股': 4783,\n",
       " '共有': 1814,\n",
       " '21': 90,\n",
       " '其中': 1841,\n",
       " '最大': 5785,\n",
       " '66': 154,\n",
       " '贵阳': 8783,\n",
       " '65': 153,\n",
       " '超过': 8873,\n",
       " '17': 51,\n",
       " '个股': 749,\n",
       " '中国': 771,\n",
       " '居前': 4042,\n",
       " '两名': 696,\n",
       " '57': 141,\n",
       " '04': 6,\n",
       " '新华': 5488,\n",
       " '券商': 2178,\n",
       " '下跌': 537,\n",
       " '近期': 9073,\n",
       " '宣布': 3868,\n",
       " '2018': 84,\n",
       " '新兴': 5483,\n",
       " '群体': 7951,\n",
       " '团队': 3055,\n",
       " '指出': 5104,\n",
       " '最新': 5790,\n",
       " '包含': 2322,\n",
       " '222': 92,\n",
       " '市值': 4175,\n",
       " '互联互通': 1065,\n",
       " '交易': 1106,\n",
       " '股票': 8051,\n",
       " '以及': 1264,\n",
       " '停牌': 1611,\n",
       " '标的': 6051,\n",
       " '由于': 7033,\n",
       " '很多': 4546,\n",
       " '权重': 5947,\n",
       " '上升': 463,\n",
       " '73': 163,\n",
       " '40': 119,\n",
       " '11': 24,\n",
       " '消费': 6547,\n",
       " '24': 94,\n",
       " '26': 98,\n",
       " '涵盖': 6564,\n",
       " '大部分': 3513,\n",
       " '动态': 2278,\n",
       " '加入': 2243,\n",
       " '增加': 3294,\n",
       " '其他': 1842,\n",
       " '尽管': 4028,\n",
       " '事件': 1018,\n",
       " '短期': 7300,\n",
       " '有所': 5820,\n",
       " '提振': 5224,\n",
       " '中长期': 860,\n",
       " '海外': 6515,\n",
       " '增量': 3308,\n",
       " '升温': 2402,\n",
       " '短期内': 7301,\n",
       " '不能': 606,\n",
       " '有效': 5821,\n",
       " '放大': 5356,\n",
       " '情景': 4713,\n",
       " '难以': 9617,\n",
       " '持续': 5097,\n",
       " '改变': 5332,\n",
       " '并未': 4302,\n",
       " '引起': 4448,\n",
       " '热情': 6755,\n",
       " '成交量': 4778,\n",
       " '较为': 8992,\n",
       " '较大': 8994,\n",
       " '变化': 2693,\n",
       " '成长': 4800,\n",
       " '全天': 1722,\n",
       " '低迷': 1391,\n",
       " '表明': 8417,\n",
       " '很少': 4549,\n",
       " '不同': 567,\n",
       " '之间': 955,\n",
       " '使得': 1457,\n",
       " '突破': 7512,\n",
       " '中期': 831,\n",
       " '依旧': 1479,\n",
       " '维持': 7840,\n",
       " '虽然': 8354,\n",
       " '至今': 8130,\n",
       " '夏普': 3351,\n",
       " '智能手机': 5737,\n",
       " '无法': 5568,\n",
       " '完全': 3776,\n",
       " '2013': 79,\n",
       " '退出': 9186,\n",
       " '月份': 5800,\n",
       " '官方': 3791,\n",
       " '回归': 3031,\n",
       " '预示': 9741,\n",
       " '很快': 4551,\n",
       " '登场': 7127,\n",
       " '登陆': 7132,\n",
       " '怎么样': 4633,\n",
       " '近日': 9072,\n",
       " '型号': 3221,\n",
       " '神秘': 7378,\n",
       " '悄然': 4695,\n",
       " '相关': 7211,\n",
       " '并非': 4305,\n",
       " '定位': 3798,\n",
       " '配备': 9333,\n",
       " '最受': 5782,\n",
       " '芯片': 8212,\n",
       " '之一': 930,\n",
       " '采用': 9343,\n",
       " '14': 43,\n",
       " '工艺': 4135,\n",
       " '设计': 8591,\n",
       " '当前': 4492,\n",
       " '只有': 2729,\n",
       " '上市': 474,\n",
       " '销售': 9440,\n",
       " '改进': 5340,\n",
       " '明显': 5665,\n",
       " '所以': 4859,\n",
       " '放在': 5355,\n",
       " '变成': 2697,\n",
       " '高端': 9945,\n",
       " '不过': 615,\n",
       " '签署': 7621,\n",
       " '协议': 2434,\n",
       " '时间': 5649,\n",
       " '既然': 5582,\n",
       " '测试': 6490,\n",
       " '说明': 8666,\n",
       " '只要': 2731,\n",
       " '时期': 5636,\n",
       " '新品': 5494,\n",
       " '之前': 938,\n",
       " '曝光': 5750,\n",
       " '竞争': 7528,\n",
       " '优势': 1316,\n",
       " '全面': 1746,\n",
       " '推出': 5197,\n",
       " '全球': 1732,\n",
       " '多达': 3421,\n",
       " '28': 101,\n",
       " '媒体': 3694,\n",
       " '沟通': 6369,\n",
       " '会上': 1330,\n",
       " '各位': 2777,\n",
       " '时代': 5626,\n",
       " '成为': 4776,\n",
       " '沙漠': 6365,\n",
       " '有利': 5811,\n",
       " '消化': 6542,\n",
       " '开盘': 4428,\n",
       " '上证': 502,\n",
       " '蓝筹': 8339,\n",
       " '带动': 4213,\n",
       " '一度': 304,\n",
       " '站上': 7526,\n",
       " '60': 145,\n",
       " '120': 36,\n",
       " '均线': 3201,\n",
       " '万达': 420,\n",
       " '电影': 7059,\n",
       " '暴跌': 5747,\n",
       " '拖累': 5070,\n",
       " '绝大多数': 7820,\n",
       " '大幅': 3467,\n",
       " '各大': 2781,\n",
       " '全线': 1737,\n",
       " '不见': 611,\n",
       " '不可': 564,\n",
       " '短线': 7304,\n",
       " '策略': 7612,\n",
       " '题材': 9771,\n",
       " '为主': 873,\n",
       " '区域': 2365,\n",
       " '概念股': 6130,\n",
       " '调整': 8685,\n",
       " '压力': 2531,\n",
       " '仍然': 1225,\n",
       " '突出': 7506,\n",
       " '再次': 1886,\n",
       " '大跌': 3507,\n",
       " '处于': 3323,\n",
       " '底部': 4348,\n",
       " '低位': 1386,\n",
       " '反复': 2606,\n",
       " '夯实': 3563,\n",
       " '蓝筹股': 8340,\n",
       " '承压': 4953,\n",
       " '后期': 2862,\n",
       " '需要': 9658,\n",
       " '接力': 5178,\n",
       " '走向': 8835,\n",
       " '今日': 1221,\n",
       " '要闻': 8471,\n",
       " '央行': 3561,\n",
       " '上海': 486,\n",
       " '总部': 4671,\n",
       " '债券': 1582,\n",
       " '北向': 2345,\n",
       " '境外': 3287,\n",
       " '投资者': 4994,\n",
       " '准入': 1957,\n",
       " '备案': 3339,\n",
       " '业务': 661,\n",
       " '指引': 5110,\n",
       " '我国': 4804,\n",
       " '信用': 1553,\n",
       " '法规': 6428,\n",
       " '标准': 6042,\n",
       " '研究': 7320,\n",
       " '加快': 2252,\n",
       " '推进': 5207,\n",
       " '房地产': 4851,\n",
       " '研究院': 7326,\n",
       " '房贷利率': 4856,\n",
       " '楼市': 6126,\n",
       " '报告': 5018,\n",
       " '许多': 8578,\n",
       " '城市': 3231,\n",
       " '个人': 739,\n",
       " '住房贷款': 1395,\n",
       " '利率': 2143,\n",
       " '程度': 7468,\n",
       " '水污染': 6316,\n",
       " '修正案': 1562,\n",
       " '草案': 8260,\n",
       " '二审': 1038,\n",
       " '更大': 5756,\n",
       " '力度': 2222,\n",
       " '保护': 1524,\n",
       " '环境': 6909,\n",
       " '结算': 7813,\n",
       " '开展': 4415,\n",
       " '一人': 254,\n",
       " '多户': 3404,\n",
       " '通知': 9238,\n",
       " '启动': 2883,\n",
       " '同一': 2821,\n",
       " '以上': 1256,\n",
       " '账户': 8749,\n",
       " '文件': 5442,\n",
       " '要求': 8469,\n",
       " '23': 93,\n",
       " '日后': 5589,\n",
       " '交易日': 1111,\n",
       " '确认': 7343,\n",
       " '十三': 2383,\n",
       " '金融市场': 9404,\n",
       " '交易额': 1113,\n",
       " '2000': 66,\n",
       " '万亿': 393,\n",
       " '宏观': 3785,\n",
       " '稳定': 7479,\n",
       " '双重': 2604,\n",
       " '因素': 3049,\n",
       " '鼓舞': 9994,\n",
       " '国际': 3117,\n",
       " '基金': 3266,\n",
       " '经理': 7789,\n",
       " '2017': 83,\n",
       " '下半年': 516,\n",
       " '股市': 8047,\n",
       " '有望': 5824,\n",
       " '十年': 2392,\n",
       " '增长': 3309,\n",
       " '创新': 2115,\n",
       " '驱动': 9879,\n",
       " '回顾': 3043,\n",
       " '早盘': 5625,\n",
       " '冲高': 1944,\n",
       " '回落': 3039,\n",
       " '走势': 8834,\n",
       " '盘面': 7177,\n",
       " '300': 104,\n",
       " '盘中': 7175,\n",
       " '新高': 5531,\n",
       " '分化': 2047,\n",
       " '相当': 7218,\n",
       " '45': 126,\n",
       " '较前': 8993,\n",
       " '76': 166,\n",
       " '跌幅': 8893,\n",
       " '101': 15,\n",
       " '97': 192,\n",
       " '98': 193,\n",
       " '创业板': 2105,\n",
       " '38': 116,\n",
       " '400': 120,\n",
       " '品种': 2939,\n",
       " '计算': 8545,\n",
       " 'st': 237,\n",
       " '涨停': 6562,\n",
       " '不足': 614,\n",
       " '亿元': 1202,\n",
       " '800': 172,\n",
       " '数据': 5424,\n",
       " '主力': 888,\n",
       " '流出': 6475,\n",
       " '一日': 317,\n",
       " '96': 191,\n",
       " '95': 190,\n",
       " '27': 99,\n",
       " '61': 148,\n",
       " '抛售': 5002,\n",
       " '13': 40,\n",
       " '介入': 1223,\n",
       " '医药': 2381,\n",
       " '制造': 2173,\n",
       " '雄安': 9625,\n",
       " '新区': 5487,\n",
       " '更是': 5761,\n",
       " '高达': 9954,\n",
       " '41': 122,\n",
       " '69': 157,\n",
       " '技术': 4963,\n",
       " '临近': 870,\n",
       " '半个': 2409,\n",
       " '小时': 3969,\n",
       " '高位': 9910,\n",
       " '留下': 7094,\n",
       " '以来': 1268,\n",
       " '每次': 6241,\n",
       " '回调': 3040,\n",
       " '相对': 7216,\n",
       " '位置': 1383,\n",
       " '即使': 2512,\n",
       " '乏力': 962,\n",
       " '总体': 4647,\n",
       " '打破': 4906,\n",
       " '攻克': 5345,\n",
       " '当日': 4501,\n",
       " '上方': 479,\n",
       " '触及': 8526,\n",
       " '无疑': 5570,\n",
       " '心理': 4599,\n",
       " '负面影响': 8729,\n",
       " '分钟': 2077,\n",
       " '级别': 7718,\n",
       " '形成': 4514,\n",
       " '一旦': 318,\n",
       " '跌破': 8894,\n",
       " '意味着': 4737,\n",
       " '如此': 3660,\n",
       " '引发': 4443,\n",
       " '综合': 7845,\n",
       " '处在': 3326,\n",
       " '区间': 2367,\n",
       " '脆弱': 8076,\n",
       " '心态': 4595,\n",
       " '碎片': 7345,\n",
       " '主线': 908,\n",
       " '模糊': 6142,\n",
       " '热点': 6760,\n",
       " '赚钱': 8810,\n",
       " '效应': 5389,\n",
       " '以下': 1257,\n",
       " '首先': 9824,\n",
       " '长期': 9468,\n",
       " '人气': 1193,\n",
       " '最近': 5795,\n",
       " '一段时间': 335,\n",
       " '整体': 5433,\n",
       " '走高': 8847,\n",
       " '基础': 3260,\n",
       " '弹性': 4466,\n",
       " '制约': 2169,\n",
       " '绝对': 7822,\n",
       " '大面积': 3521,\n",
       " '与此同时': 621,\n",
       " '现在': 6926,\n",
       " '非常': 9683,\n",
       " '频繁': 9767,\n",
       " '今天': 1216,\n",
       " '明天': 5661,\n",
       " '就要': 4009,\n",
       " '打击': 4896,\n",
       " '积极性': 7452,\n",
       " '其次': 1847,\n",
       " '反弹': 2612,\n",
       " '重要': 9378,\n",
       " '拉升': 5047,\n",
       " '这种': 9111,\n",
       " '大涨': 3482,\n",
       " '似乎': 1378,\n",
       " '一直': 343,\n",
       " '吸引': 2888,\n",
       " '加码': 2264,\n",
       " '中小': 808,\n",
       " '面临': 9688,\n",
       " '强大': 4475,\n",
       " '午后': 2407,\n",
       " '复星': 3344,\n",
       " '收复': 5315,\n",
       " '投资': 4992,\n",
       " '建议': 4393,\n",
       " '始终': 3666,\n",
       " '缺乏': 7880,\n",
       " '支持': 5297,\n",
       " '背景': 8060,\n",
       " '后续': 2865,\n",
       " '跟上': 8901,\n",
       " '活跃': 6467,\n",
       " '降低': 9564,\n",
       " '配置': 9335,\n",
       " '加大': 2246,\n",
       " '纠缠': 7701,\n",
       " '促使': 1505,\n",
       " '能否': 8071,\n",
       " '真正': 7262,\n",
       " '观察': 8482,\n",
       " '基于': 3244,\n",
       " '组合': 7751,\n",
       " '因此': 3047,\n",
       " '股指': 8048,\n",
       " '下方': 523,\n",
       " '获得': 8282,\n",
       " '操作': 5288,\n",
       " '粤港澳': 7659,\n",
       " '大湾': 3483,\n",
       " '论坛': 8580,\n",
       " '腾讯': 8088,\n",
       " '人工智能': 1178,\n",
       " '百度': 7150,\n",
       " '举办': 918,\n",
       " 'ai': 196,\n",
       " '开发者': 4408,\n",
       " '大会': 3430,\n",
       " '亚马逊': 1093,\n",
       " '代表': 1251,\n",
       " '互联网': 1066,\n",
       " '巨头': 4144,\n",
       " '一致': 358,\n",
       " '看好': 7245,\n",
       " '发展': 2629,\n",
       " '产业': 1126,\n",
       " '叠加': 2704,\n",
       " '政策': 5378,\n",
       " '扶持': 4942,\n",
       " '景气': 5726,\n",
       " '一个月': 249,\n",
       " '三次': 448,\n",
       " '加之': 2239,\n",
       " '下周': 520,\n",
       " '月末': 5804,\n",
       " '值得': 1588,\n",
       " '做好': 1603,\n",
       " '应对': 4339,\n",
       " '月初': 5801,\n",
       " '为首': 881,\n",
       " '模式': 6139,\n",
       " '保持': 1527,\n",
       " '随着': 9606,\n",
       " '带来': 4218,\n",
       " '刷新': 2177,\n",
       " '高点': 9942,\n",
       " '降温': 9568,\n",
       " '加剧': 2245,\n",
       " '现象': 6936,\n",
       " '必须': 4607,\n",
       " '重视': 9381,\n",
       " '红利': 7702,\n",
       " '军工': 1896,\n",
       " '上周': 468,\n",
       " '王者': 6898,\n",
       " '荣耀': 8263,\n",
       " '类似': 7655,\n",
       " '来自': 5992,\n",
       " '比较': 6259,\n",
       " '敏感': 5395,\n",
       " '半年': 2414,\n",
       " '容易': 3894,\n",
       " '实际上': 3835,\n",
       " '美联储': 7944,\n",
       " '加息': 2253,\n",
       " '缩表': 7875,\n",
       " '行为': 8377,\n",
       " '货币政策': 8751,\n",
       " '开启': 4409,\n",
       " '流行': 6482,\n",
       " '很难': 4553,\n",
       " '中性': 821,\n",
       " '偏紧': 1600,\n",
       " '常态': 4237,\n",
       " '基本面': 3257,\n",
       " '阶段': 9522,\n",
       " '一段': 334,\n",
       " '高速': 9956,\n",
       " '增速': 3307,\n",
       " '下滑': 530,\n",
       " '而是': 7987,\n",
       " '周期': 2910,\n",
       " '适度': 9196,\n",
       " '供给': 1475,\n",
       " '结构性': 7809,\n",
       " '改革': 5342,\n",
       " '决定': 1946,\n",
       " '运行': 9062,\n",
       " '之下': 932,\n",
       " '前期': 2194,\n",
       " '单纯': 2448,\n",
       " '依托': 1477,\n",
       " '能够': 8072,\n",
       " '担忧': 5043,\n",
       " '尤其': 3998,\n",
       " '越来越': 8876,\n",
       " '甚至': 6986,\n",
       " '排除': 5168,\n",
       " '选择': 9210,\n",
       " '伴随': 1377,\n",
       " '谨慎': 8709,\n",
       " '气氛': 6296,\n",
       " '对于': 3910,\n",
       " '梳理': 6104,\n",
       " '警惕': 8536,\n",
       " '兑现': 1681,\n",
       " '利润': 2140,\n",
       " '重新': 9372,\n",
       " '机会': 5924,\n",
       " '阻力': 9524,\n",
       " '密集': 3905,\n",
       " '及时': 2587,\n",
       " '减仓': 1964,\n",
       " '毕竟': 6265,\n",
       " '通道': 9250,\n",
       " '扭转': 4938,\n",
       " '暂时': 5740,\n",
       " '持有': 5095,\n",
       " '主板': 903,\n",
       " '终于': 7769,\n",
       " '遗憾': 9283,\n",
       " '之际': 956,\n",
       " '周五': 2903,\n",
       " '怎样': 4634,\n",
       " '登记': 7131,\n",
       " '托管': 4914,\n",
       " '实施细则': 3825,\n",
       " '实施': 3823,\n",
       " '发改委': 2642,\n",
       " '抓紧': 4976,\n",
       " '制定': 2162,\n",
       " '出台': 1999,\n",
       " '价格': 1279,\n",
       " '监管': 7171,\n",
       " '规则': 8492,\n",
       " '社部': 7369,\n",
       " '月底': 5802,\n",
       " '基本': 3254,\n",
       " '实现': 3828,\n",
       " '跨省': 8909,\n",
       " '异地': 4440,\n",
       " '就医': 4004,\n",
       " '直接': 7200,\n",
       " '成品油': 4788,\n",
       " '年内': 4277,\n",
       " '下调': 536,\n",
       " '降幅': 9566,\n",
       " '国防': 3114,\n",
       " '科工': 7421,\n",
       " '混改': 6594,\n",
       " '资产': 8789,\n",
       " '证券化': 8599,\n",
       " '进程': 9139,\n",
       " '广电总局': 4327,\n",
       " '新浪': 5512,\n",
       " '微博': 4576,\n",
       " '网站': 7894,\n",
       " '关停': 1820,\n",
       " '视听': 8503,\n",
       " '节目': 8200,\n",
       " '欧美': 6172,\n",
       " '稳健': 7478,\n",
       " '国内': 3075,\n",
       " '形式': 4512,\n",
       " '整理': 5437,\n",
       " '1800': 55,\n",
       " '相比': 7222,\n",
       " '显得': 5699,\n",
       " '弱势': 4465,\n",
       " '转换': 8971,\n",
       " '随时': 9604,\n",
       " '第一': 7548,\n",
       " '第二': 7576,\n",
       " '实质性': 3832,\n",
       " '低开': 1387,\n",
       " '可能性': 2742,\n",
       " '防止': 9511,\n",
       " '极为': 6002,\n",
       " ...}"
      ]
     },
     "execution_count": 148,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vectorized.vocabulary_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,\n",
       "        0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], dtype=int64),\n",
       " array([  35,   46,  302,  912,  933, 1546, 1743, 1819, 1844, 2558, 2635,\n",
       "        2734, 3974, 4113, 4132, 4405, 4505, 4525, 4802, 4878, 5628, 5738,\n",
       "        5760, 5808, 5892, 6195, 6880, 7319, 7338, 7607, 7662, 9082, 9141,\n",
       "        9631], dtype=int64))"
      ]
     },
     "execution_count": 149,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.where(X[0].toarray()) # get the positions which values are not zero"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'此外   自 本周   6 月 12 日   起   除 小米 手机 6 等 15 款 机型 外   其余 机型 已 暂停 更新 发布   含 开发 版   体验版 内测   稳定版 暂不受 影响   以 确保 工程师 可以 集中 全部 精力 进行 系统优化 工作   有人 猜测 这 也 是 将 精力 主要 用到 MIUI   9 的 研发 之中   MIUI   8 去年 5 月 发布   距今已有 一年 有余   也 是 时候 更新换代 了   当然   关于 MIUI   9 的 确切 信息   我们 还是 等待 官方消息'"
      ]
     },
     "execution_count": 150,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_content[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'此外': 6195,\n",
       " '本周': 5892,\n",
       " '12': 35,\n",
       " '小米': 3974,\n",
       " '手机': 4878,\n",
       " '15': 46,\n",
       " '其余': 1844,\n",
       " '暂停': 5738,\n",
       " '更新': 5760,\n",
       " '发布': 2635,\n",
       " '开发': 4405,\n",
       " '影响': 4525,\n",
       " '确保': 7338,\n",
       " '工程师': 4132,\n",
       " '可以': 2734,\n",
       " '集中': 9631,\n",
       " '全部': 1743,\n",
       " '精力': 7662,\n",
       " '进行': 9141,\n",
       " '工作': 4113,\n",
       " '有人': 5808,\n",
       " '猜测': 6880,\n",
       " '主要': 912,\n",
       " '研发': 7319,\n",
       " '之中': 933,\n",
       " '去年': 2558,\n",
       " '一年': 302,\n",
       " '时候': 5628,\n",
       " '当然': 4505,\n",
       " '关于': 1819,\n",
       " '信息': 1546,\n",
       " '我们': 4802,\n",
       " '还是': 9082,\n",
       " '等待': 7607,\n",
       " '作为': 1440,\n",
       " '唯一': 2966,\n",
       " '通过': 9249,\n",
       " '10': 12,\n",
       " '平台': 4255,\n",
       " '认证': 8554,\n",
       " '处理器': 3331,\n",
       " '强调': 4481,\n",
       " '不会': 551,\n",
       " '因为': 3045,\n",
       " '考虑': 7980,\n",
       " '性能': 4642,\n",
       " '核心': 6069,\n",
       " '相反': 7212,\n",
       " '他们': 1242,\n",
       " '联手': 8020,\n",
       " '微软': 4579,\n",
       " '找到': 4949,\n",
       " '一种': 344,\n",
       " '适合': 9193,\n",
       " '兼顾': 1864,\n",
       " '完美': 3783,\n",
       " '方案': 5537,\n",
       " '报道': 5026,\n",
       " '已经': 4156,\n",
       " '拿到': 5092,\n",
       " '一些': 253,\n",
       " '以便': 1259,\n",
       " '更好': 5757,\n",
       " '理解': 6966,\n",
       " '架构': 6022,\n",
       " '资料': 8791,\n",
       " '显示': 5703,\n",
       " '一款': 332,\n",
       " '集成': 9638,\n",
       " '传统': 1355,\n",
       " '节省': 8201,\n",
       " '至少': 8132,\n",
       " '30': 103,\n",
       " '空间': 7498,\n",
       " '按计划': 5124,\n",
       " '今年': 1217,\n",
       " '联想': 8019,\n",
       " '首发': 9825,\n",
       " '电脑': 7067,\n",
       " '预计': 9745,\n",
       " '形态': 4513,\n",
       " '产品': 1135,\n",
       " '只是': 2728,\n",
       " '开始': 4412,\n",
       " '未来': 5886,\n",
       " '也许': 982,\n",
       " '见到': 8475,\n",
       " '三星': 443,\n",
       " '华为': 2417,\n",
       " '澎湃': 6699,\n",
       " '进入': 9123,\n",
       " '此前': 6193,\n",
       " '搭载': 5252,\n",
       " '电池': 7063,\n",
       " '规格': 8495,\n",
       " '至于': 8129,\n",
       " '可能': 2741,\n",
       " '所说': 4870,\n",
       " '设定': 8585,\n",
       " '有关': 5810,\n",
       " '按照': 5122,\n",
       " '目前': 7188,\n",
       " '掌握': 5160,\n",
       " '拥有': 5083,\n",
       " '摄像头': 5256,\n",
       " '惊喜': 4717,\n",
       " '根据': 6078,\n",
       " '京东': 1144,\n",
       " '泄露': 6406,\n",
       " '售价': 2964,\n",
       " '应该': 4344,\n",
       " '某个': 6025,\n",
       " '这是': 9105,\n",
       " '18': 53,\n",
       " '葡萄牙': 8322,\n",
       " '中部': 858,\n",
       " '地区': 3168,\n",
       " '拍摄': 5064,\n",
       " '森林': 6119,\n",
       " '大火': 3485,\n",
       " '汽车': 6355,\n",
       " '新华社': 5489,\n",
       " '记者': 8569,\n",
       " '标题': 6055,\n",
       " '44': 125,\n",
       " '女子': 3625,\n",
       " '深圳': 6576,\n",
       " '网友': 7888,\n",
       " '暴雨': 5748,\n",
       " '交警': 1120,\n",
       " '昨日': 5694,\n",
       " '清晨': 6604,\n",
       " '发现': 2647,\n",
       " '行走': 8392,\n",
       " '快速': 4622,\n",
       " '期间': 5881,\n",
       " '赶紧': 8852,\n",
       " '一路': 365,\n",
       " '那么': 9299,\n",
       " '事发': 1024,\n",
       " '到底': 2154,\n",
       " '发生': 2649,\n",
       " '什么': 1208,\n",
       " '一起': 364,\n",
       " '还原': 9081,\n",
       " '现场': 6927,\n",
       " '大队': 3517,\n",
       " '中队': 863,\n",
       " '女生': 3631,\n",
       " '一位': 259,\n",
       " '高大': 9919,\n",
       " '说话': 8668,\n",
       " '90': 184,\n",
       " '青年': 9672,\n",
       " '介绍': 1224,\n",
       " '16': 49,\n",
       " '早上': 5617,\n",
       " '36': 113,\n",
       " '正在': 6180,\n",
       " '附近': 9550,\n",
       " '接到': 5177,\n",
       " '机动车': 5928,\n",
       " '危险': 2511,\n",
       " '随后': 9602,\n",
       " '寻找': 3932,\n",
       " '大概': 3476,\n",
       " '大道': 3511,\n",
       " '出口': 1998,\n",
       " '方向': 5535,\n",
       " '身上': 8929,\n",
       " '另外': 2722,\n",
       " '一名': 276,\n",
       " '情绪': 4714,\n",
       " '尝试': 3996,\n",
       " '交流': 1115,\n",
       " '离开': 7403,\n",
       " '愿意': 4764,\n",
       " '接受': 5179,\n",
       " '继续': 7832,\n",
       " '缓慢': 7862,\n",
       " '此时': 6197,\n",
       " '路边': 8918,\n",
       " '聚集': 8034,\n",
       " '市民': 4192,\n",
       " '围观': 3065,\n",
       " '为了': 874,\n",
       " '刺激': 2179,\n",
       " '一边': 370,\n",
       " '群众': 7950,\n",
       " '警方': 8538,\n",
       " '提供': 5215,\n",
       " '一份': 257,\n",
       " '视频': 8509,\n",
       " '了解': 1007,\n",
       " '25': 95,\n",
       " '出现': 2021,\n",
       " '监控': 7165,\n",
       " '穿着': 7502,\n",
       " '白色': 7140,\n",
       " '沿着': 6403,\n",
       " '当时': 4502,\n",
       " '正值': 6179,\n",
       " '上班': 495,\n",
       " '高峰期': 9927,\n",
       " '路上': 8911,\n",
       " '纷纷': 7737,\n",
       " '驻足': 9885,\n",
       " '观望': 8485,\n",
       " '不少': 576,\n",
       " '车辆': 8955,\n",
       " '速度': 9253,\n",
       " '脚步': 8077,\n",
       " '依然': 1483,\n",
       " '行进': 8393,\n",
       " '中间': 861,\n",
       " '一辆': 368,\n",
       " '货车': 8755,\n",
       " '镜头': 9454,\n",
       " '但是': 1379,\n",
       " '全身': 1740,\n",
       " '朝着': 5865,\n",
       " '周围': 2907,\n",
       " '没有': 6373,\n",
       " '或者': 4811,\n",
       " '上前': 462,\n",
       " '看到': 7244,\n",
       " '这样': 9107,\n",
       " '情况': 4708,\n",
       " '恐怕': 4678,\n",
       " '大家': 3459,\n",
       " '办法': 2232,\n",
       " '面对': 9693,\n",
       " '这一': 9085,\n",
       " '表示': 8421,\n",
       " '根本': 6079,\n",
       " '不敢': 588,\n",
       " '心里': 4603,\n",
       " '感觉': 4760,\n",
       " '尴尬': 4014,\n",
       " '跟随': 8905,\n",
       " '作出': 1441,\n",
       " '举动': 919,\n",
       " '突然': 7511,\n",
       " '靠近': 9687,\n",
       " '上面': 510,\n",
       " '控制': 5190,\n",
       " '远离': 9146,\n",
       " '衣服': 8404,\n",
       " '只能': 2730,\n",
       " '紧紧': 7691,\n",
       " '一只': 272,\n",
       " '后面': 2867,\n",
       " '耐心': 7989,\n",
       " '安慰': 3766,\n",
       " '听到': 2879,\n",
       " '不断': 589,\n",
       " '重复': 9364,\n",
       " '一句': 271,\n",
       " '要是': 8468,\n",
       " '遭遇': 9288,\n",
       " '不时': 593,\n",
       " '试图': 8626,\n",
       " '大暴雨': 3472,\n",
       " '眼睛': 7270,\n",
       " '瞬间': 7288,\n",
       " '帮助': 4224,\n",
       " '大雨': 3519,\n",
       " '来说': 5995,\n",
       " '回家': 3029,\n",
       " '然后': 6774,\n",
       " '照片': 6781,\n",
       " '旁边': 5546,\n",
       " '场景': 3196,\n",
       " '下来': 525,\n",
       " '带到': 4212,\n",
       " '派出所': 6470,\n",
       " '姑娘': 3669,\n",
       " '什么样': 1209,\n",
       " '事情': 1029,\n",
       " '透露': 9219,\n",
       " '湖北': 6645,\n",
       " '家属': 3883,\n",
       " '反映': 2615,\n",
       " '三天': 437,\n",
       " '陈某': 9557,\n",
       " '老家': 7963,\n",
       " '导致': 3939,\n",
       " '异常': 4441,\n",
       " '产生': 1139,\n",
       " '送往': 9191,\n",
       " '医院': 2382,\n",
       " '治疗': 6400,\n",
       " '大大': 3452,\n",
       " '希望': 4207,\n",
       " '康复': 4363,\n",
       " '其实': 1846,\n",
       " '到来': 2156,\n",
       " '存在': 3707,\n",
       " '年龄': 4296,\n",
       " '限制': 9577,\n",
       " '你们': 1451,\n",
       " '原因': 2542,\n",
       " '这个': 9086,\n",
       " '哥哥': 2953,\n",
       " '一辈子': 369,\n",
       " '自己': 8099,\n",
       " '家人': 3878,\n",
       " '同时': 2831,\n",
       " '感谢': 4761,\n",
       " '聆听': 7996,\n",
       " '心灵': 4598,\n",
       " '点赞': 6745,\n",
       " '警察': 8534,\n",
       " '就是': 4005,\n",
       " '男子': 7080,\n",
       " '20': 64,\n",
       " '50': 132,\n",
       " '上午': 464,\n",
       " '公安局': 1771,\n",
       " '一个': 247,\n",
       " '报警': 5025,\n",
       " '电话': 7072,\n",
       " '声称': 3321,\n",
       " '遭到': 9286,\n",
       " '侵害': 1492,\n",
       " '几个': 1975,\n",
       " '关键词': 1834,\n",
       " '民警': 6287,\n",
       " '紧张': 7688,\n",
       " '起来': 8855,\n",
       " '受到': 2677,\n",
       " 'a股': 198,\n",
       " '纳入': 7730,\n",
       " 'msci': 219,\n",
       " '指数': 5115,\n",
       " '利好': 2136,\n",
       " '消息': 6544,\n",
       " '市场': 4178,\n",
       " '周三': 2901,\n",
       " '再度': 1884,\n",
       " '上演': 494,\n",
       " '行情': 8383,\n",
       " '周四': 2906,\n",
       " '金融股': 9409,\n",
       " '白马股': 7143,\n",
       " '表现': 8420,\n",
       " '尾盘': 4030,\n",
       " '跳水': 8922,\n",
       " '之后': 940,\n",
       " '金融': 9400,\n",
       " '板块': 6001,\n",
       " '状态': 6865,\n",
       " '分析': 2060,\n",
       " '人士': 1172,\n",
       " '认为': 8548,\n",
       " '受益': 2685,\n",
       " '估值': 1374,\n",
       " '资金': 8803,\n",
       " '青睐': 9679,\n",
       " '存量': 3711,\n",
       " '博弈': 2476,\n",
       " '格局': 6082,\n",
       " '风格': 9784,\n",
       " '震荡': 9662,\n",
       " '延续': 4373,\n",
       " '流动性': 6477,\n",
       " '改善': 5333,\n",
       " '经济': 7779,\n",
       " '预期': 9738,\n",
       " '有助于': 5814,\n",
       " '支撑': 5301,\n",
       " '大盘': 3490,\n",
       " '逐步': 9222,\n",
       " '再现': 1887,\n",
       " '未能': 5889,\n",
       " '上行': 501,\n",
       " '态势': 4629,\n",
       " '两市': 710,\n",
       " '成交': 4777,\n",
       " '小幅': 3964,\n",
       " '放量': 5367,\n",
       " '29': 102,\n",
       " '中信': 757,\n",
       " '一级': 352,\n",
       " '行业': 8375,\n",
       " '收盘': 5320,\n",
       " '银行': 9434,\n",
       " '两个': 688,\n",
       " '分别': 2045,\n",
       " '上涨': 490,\n",
       " '80': 171,\n",
       " '二级': 1048,\n",
       " '来看': 5991,\n",
       " '股份制': 8046,\n",
       " '涨幅': 6563,\n",
       " '最高': 5797,\n",
       " '达到': 9021,\n",
       " '22': 91,\n",
       " '82': 175,\n",
       " '信托': 1552,\n",
       " '及其': 2586,\n",
       " '64': 152,\n",
       " '保险': 1536,\n",
       " '34': 109,\n",
       " '证券': 8598,\n",
       " '06': 8,\n",
       " '成分股': 4783,\n",
       " '共有': 1814,\n",
       " '21': 90,\n",
       " '其中': 1841,\n",
       " '最大': 5785,\n",
       " '66': 154,\n",
       " '贵阳': 8783,\n",
       " '65': 153,\n",
       " '超过': 8873,\n",
       " '17': 51,\n",
       " '个股': 749,\n",
       " '中国': 771,\n",
       " '居前': 4042,\n",
       " '两名': 696,\n",
       " '57': 141,\n",
       " '04': 6,\n",
       " '新华': 5488,\n",
       " '券商': 2178,\n",
       " '下跌': 537,\n",
       " '近期': 9073,\n",
       " '宣布': 3868,\n",
       " '2018': 84,\n",
       " '新兴': 5483,\n",
       " '群体': 7951,\n",
       " '团队': 3055,\n",
       " '指出': 5104,\n",
       " '最新': 5790,\n",
       " '包含': 2322,\n",
       " '222': 92,\n",
       " '市值': 4175,\n",
       " '互联互通': 1065,\n",
       " '交易': 1106,\n",
       " '股票': 8051,\n",
       " '以及': 1264,\n",
       " '停牌': 1611,\n",
       " '标的': 6051,\n",
       " '由于': 7033,\n",
       " '很多': 4546,\n",
       " '权重': 5947,\n",
       " '上升': 463,\n",
       " '73': 163,\n",
       " '40': 119,\n",
       " '11': 24,\n",
       " '消费': 6547,\n",
       " '24': 94,\n",
       " '26': 98,\n",
       " '涵盖': 6564,\n",
       " '大部分': 3513,\n",
       " '动态': 2278,\n",
       " '加入': 2243,\n",
       " '增加': 3294,\n",
       " '其他': 1842,\n",
       " '尽管': 4028,\n",
       " '事件': 1018,\n",
       " '短期': 7300,\n",
       " '有所': 5820,\n",
       " '提振': 5224,\n",
       " '中长期': 860,\n",
       " '海外': 6515,\n",
       " '增量': 3308,\n",
       " '升温': 2402,\n",
       " '短期内': 7301,\n",
       " '不能': 606,\n",
       " '有效': 5821,\n",
       " '放大': 5356,\n",
       " '情景': 4713,\n",
       " '难以': 9617,\n",
       " '持续': 5097,\n",
       " '改变': 5332,\n",
       " '并未': 4302,\n",
       " '引起': 4448,\n",
       " '热情': 6755,\n",
       " '成交量': 4778,\n",
       " '较为': 8992,\n",
       " '较大': 8994,\n",
       " '变化': 2693,\n",
       " '成长': 4800,\n",
       " '全天': 1722,\n",
       " '低迷': 1391,\n",
       " '表明': 8417,\n",
       " '很少': 4549,\n",
       " '不同': 567,\n",
       " '之间': 955,\n",
       " '使得': 1457,\n",
       " '突破': 7512,\n",
       " '中期': 831,\n",
       " '依旧': 1479,\n",
       " '维持': 7840,\n",
       " '虽然': 8354,\n",
       " '至今': 8130,\n",
       " '夏普': 3351,\n",
       " '智能手机': 5737,\n",
       " '无法': 5568,\n",
       " '完全': 3776,\n",
       " '2013': 79,\n",
       " '退出': 9186,\n",
       " '月份': 5800,\n",
       " '官方': 3791,\n",
       " '回归': 3031,\n",
       " '预示': 9741,\n",
       " '很快': 4551,\n",
       " '登场': 7127,\n",
       " '登陆': 7132,\n",
       " '怎么样': 4633,\n",
       " '近日': 9072,\n",
       " '型号': 3221,\n",
       " '神秘': 7378,\n",
       " '悄然': 4695,\n",
       " '相关': 7211,\n",
       " '并非': 4305,\n",
       " '定位': 3798,\n",
       " '配备': 9333,\n",
       " '最受': 5782,\n",
       " '芯片': 8212,\n",
       " '之一': 930,\n",
       " '采用': 9343,\n",
       " '14': 43,\n",
       " '工艺': 4135,\n",
       " '设计': 8591,\n",
       " '当前': 4492,\n",
       " '只有': 2729,\n",
       " '上市': 474,\n",
       " '销售': 9440,\n",
       " '改进': 5340,\n",
       " '明显': 5665,\n",
       " '所以': 4859,\n",
       " '放在': 5355,\n",
       " '变成': 2697,\n",
       " '高端': 9945,\n",
       " '不过': 615,\n",
       " '签署': 7621,\n",
       " '协议': 2434,\n",
       " '时间': 5649,\n",
       " '既然': 5582,\n",
       " '测试': 6490,\n",
       " '说明': 8666,\n",
       " '只要': 2731,\n",
       " '时期': 5636,\n",
       " '新品': 5494,\n",
       " '之前': 938,\n",
       " '曝光': 5750,\n",
       " '竞争': 7528,\n",
       " '优势': 1316,\n",
       " '全面': 1746,\n",
       " '推出': 5197,\n",
       " '全球': 1732,\n",
       " '多达': 3421,\n",
       " '28': 101,\n",
       " '媒体': 3694,\n",
       " '沟通': 6369,\n",
       " '会上': 1330,\n",
       " '各位': 2777,\n",
       " '时代': 5626,\n",
       " '成为': 4776,\n",
       " '沙漠': 6365,\n",
       " '有利': 5811,\n",
       " '消化': 6542,\n",
       " '开盘': 4428,\n",
       " '上证': 502,\n",
       " '蓝筹': 8339,\n",
       " '带动': 4213,\n",
       " '一度': 304,\n",
       " '站上': 7526,\n",
       " '60': 145,\n",
       " '120': 36,\n",
       " '均线': 3201,\n",
       " '万达': 420,\n",
       " '电影': 7059,\n",
       " '暴跌': 5747,\n",
       " '拖累': 5070,\n",
       " '绝大多数': 7820,\n",
       " '大幅': 3467,\n",
       " '各大': 2781,\n",
       " '全线': 1737,\n",
       " '不见': 611,\n",
       " '不可': 564,\n",
       " '短线': 7304,\n",
       " '策略': 7612,\n",
       " '题材': 9771,\n",
       " '为主': 873,\n",
       " '区域': 2365,\n",
       " '概念股': 6130,\n",
       " '调整': 8685,\n",
       " '压力': 2531,\n",
       " '仍然': 1225,\n",
       " '突出': 7506,\n",
       " '再次': 1886,\n",
       " '大跌': 3507,\n",
       " '处于': 3323,\n",
       " '底部': 4348,\n",
       " '低位': 1386,\n",
       " '反复': 2606,\n",
       " '夯实': 3563,\n",
       " '蓝筹股': 8340,\n",
       " '承压': 4953,\n",
       " '后期': 2862,\n",
       " '需要': 9658,\n",
       " '接力': 5178,\n",
       " '走向': 8835,\n",
       " '今日': 1221,\n",
       " '要闻': 8471,\n",
       " '央行': 3561,\n",
       " '上海': 486,\n",
       " '总部': 4671,\n",
       " '债券': 1582,\n",
       " '北向': 2345,\n",
       " '境外': 3287,\n",
       " '投资者': 4994,\n",
       " '准入': 1957,\n",
       " '备案': 3339,\n",
       " '业务': 661,\n",
       " '指引': 5110,\n",
       " '我国': 4804,\n",
       " '信用': 1553,\n",
       " '法规': 6428,\n",
       " '标准': 6042,\n",
       " '研究': 7320,\n",
       " '加快': 2252,\n",
       " '推进': 5207,\n",
       " '房地产': 4851,\n",
       " '研究院': 7326,\n",
       " '房贷利率': 4856,\n",
       " '楼市': 6126,\n",
       " '报告': 5018,\n",
       " '许多': 8578,\n",
       " '城市': 3231,\n",
       " '个人': 739,\n",
       " '住房贷款': 1395,\n",
       " '利率': 2143,\n",
       " '程度': 7468,\n",
       " '水污染': 6316,\n",
       " '修正案': 1562,\n",
       " '草案': 8260,\n",
       " '二审': 1038,\n",
       " '更大': 5756,\n",
       " '力度': 2222,\n",
       " '保护': 1524,\n",
       " '环境': 6909,\n",
       " '结算': 7813,\n",
       " '开展': 4415,\n",
       " '一人': 254,\n",
       " '多户': 3404,\n",
       " '通知': 9238,\n",
       " '启动': 2883,\n",
       " '同一': 2821,\n",
       " '以上': 1256,\n",
       " '账户': 8749,\n",
       " '文件': 5442,\n",
       " '要求': 8469,\n",
       " '23': 93,\n",
       " '日后': 5589,\n",
       " '交易日': 1111,\n",
       " '确认': 7343,\n",
       " '十三': 2383,\n",
       " '金融市场': 9404,\n",
       " '交易额': 1113,\n",
       " '2000': 66,\n",
       " '万亿': 393,\n",
       " '宏观': 3785,\n",
       " '稳定': 7479,\n",
       " '双重': 2604,\n",
       " '因素': 3049,\n",
       " '鼓舞': 9994,\n",
       " '国际': 3117,\n",
       " '基金': 3266,\n",
       " '经理': 7789,\n",
       " '2017': 83,\n",
       " '下半年': 516,\n",
       " '股市': 8047,\n",
       " '有望': 5824,\n",
       " '十年': 2392,\n",
       " '增长': 3309,\n",
       " '创新': 2115,\n",
       " '驱动': 9879,\n",
       " '回顾': 3043,\n",
       " '早盘': 5625,\n",
       " '冲高': 1944,\n",
       " '回落': 3039,\n",
       " '走势': 8834,\n",
       " '盘面': 7177,\n",
       " '300': 104,\n",
       " '盘中': 7175,\n",
       " '新高': 5531,\n",
       " '分化': 2047,\n",
       " '相当': 7218,\n",
       " '45': 126,\n",
       " '较前': 8993,\n",
       " '76': 166,\n",
       " '跌幅': 8893,\n",
       " '101': 15,\n",
       " '97': 192,\n",
       " '98': 193,\n",
       " '创业板': 2105,\n",
       " '38': 116,\n",
       " '400': 120,\n",
       " '品种': 2939,\n",
       " '计算': 8545,\n",
       " 'st': 237,\n",
       " '涨停': 6562,\n",
       " '不足': 614,\n",
       " '亿元': 1202,\n",
       " '800': 172,\n",
       " '数据': 5424,\n",
       " '主力': 888,\n",
       " '流出': 6475,\n",
       " '一日': 317,\n",
       " '96': 191,\n",
       " '95': 190,\n",
       " '27': 99,\n",
       " '61': 148,\n",
       " '抛售': 5002,\n",
       " '13': 40,\n",
       " '介入': 1223,\n",
       " '医药': 2381,\n",
       " '制造': 2173,\n",
       " '雄安': 9625,\n",
       " '新区': 5487,\n",
       " '更是': 5761,\n",
       " '高达': 9954,\n",
       " '41': 122,\n",
       " '69': 157,\n",
       " '技术': 4963,\n",
       " '临近': 870,\n",
       " '半个': 2409,\n",
       " '小时': 3969,\n",
       " '高位': 9910,\n",
       " '留下': 7094,\n",
       " '以来': 1268,\n",
       " '每次': 6241,\n",
       " '回调': 3040,\n",
       " '相对': 7216,\n",
       " '位置': 1383,\n",
       " '即使': 2512,\n",
       " '乏力': 962,\n",
       " '总体': 4647,\n",
       " '打破': 4906,\n",
       " '攻克': 5345,\n",
       " '当日': 4501,\n",
       " '上方': 479,\n",
       " '触及': 8526,\n",
       " '无疑': 5570,\n",
       " '心理': 4599,\n",
       " '负面影响': 8729,\n",
       " '分钟': 2077,\n",
       " '级别': 7718,\n",
       " '形成': 4514,\n",
       " '一旦': 318,\n",
       " '跌破': 8894,\n",
       " '意味着': 4737,\n",
       " '如此': 3660,\n",
       " '引发': 4443,\n",
       " '综合': 7845,\n",
       " '处在': 3326,\n",
       " '区间': 2367,\n",
       " '脆弱': 8076,\n",
       " '心态': 4595,\n",
       " '碎片': 7345,\n",
       " '主线': 908,\n",
       " '模糊': 6142,\n",
       " '热点': 6760,\n",
       " '赚钱': 8810,\n",
       " '效应': 5389,\n",
       " '以下': 1257,\n",
       " '首先': 9824,\n",
       " '长期': 9468,\n",
       " '人气': 1193,\n",
       " '最近': 5795,\n",
       " '一段时间': 335,\n",
       " '整体': 5433,\n",
       " '走高': 8847,\n",
       " '基础': 3260,\n",
       " '弹性': 4466,\n",
       " '制约': 2169,\n",
       " '绝对': 7822,\n",
       " '大面积': 3521,\n",
       " '与此同时': 621,\n",
       " '现在': 6926,\n",
       " '非常': 9683,\n",
       " '频繁': 9767,\n",
       " '今天': 1216,\n",
       " '明天': 5661,\n",
       " '就要': 4009,\n",
       " '打击': 4896,\n",
       " '积极性': 7452,\n",
       " '其次': 1847,\n",
       " '反弹': 2612,\n",
       " '重要': 9378,\n",
       " '拉升': 5047,\n",
       " '这种': 9111,\n",
       " '大涨': 3482,\n",
       " '似乎': 1378,\n",
       " '一直': 343,\n",
       " '吸引': 2888,\n",
       " '加码': 2264,\n",
       " '中小': 808,\n",
       " '面临': 9688,\n",
       " '强大': 4475,\n",
       " '午后': 2407,\n",
       " '复星': 3344,\n",
       " '收复': 5315,\n",
       " '投资': 4992,\n",
       " '建议': 4393,\n",
       " '始终': 3666,\n",
       " '缺乏': 7880,\n",
       " '支持': 5297,\n",
       " '背景': 8060,\n",
       " '后续': 2865,\n",
       " '跟上': 8901,\n",
       " '活跃': 6467,\n",
       " '降低': 9564,\n",
       " '配置': 9335,\n",
       " '加大': 2246,\n",
       " '纠缠': 7701,\n",
       " '促使': 1505,\n",
       " '能否': 8071,\n",
       " '真正': 7262,\n",
       " '观察': 8482,\n",
       " '基于': 3244,\n",
       " '组合': 7751,\n",
       " '因此': 3047,\n",
       " '股指': 8048,\n",
       " '下方': 523,\n",
       " '获得': 8282,\n",
       " '操作': 5288,\n",
       " '粤港澳': 7659,\n",
       " '大湾': 3483,\n",
       " '论坛': 8580,\n",
       " '腾讯': 8088,\n",
       " '人工智能': 1178,\n",
       " '百度': 7150,\n",
       " '举办': 918,\n",
       " 'ai': 196,\n",
       " '开发者': 4408,\n",
       " '大会': 3430,\n",
       " '亚马逊': 1093,\n",
       " '代表': 1251,\n",
       " '互联网': 1066,\n",
       " '巨头': 4144,\n",
       " '一致': 358,\n",
       " '看好': 7245,\n",
       " '发展': 2629,\n",
       " '产业': 1126,\n",
       " '叠加': 2704,\n",
       " '政策': 5378,\n",
       " '扶持': 4942,\n",
       " '景气': 5726,\n",
       " '一个月': 249,\n",
       " '三次': 448,\n",
       " '加之': 2239,\n",
       " '下周': 520,\n",
       " '月末': 5804,\n",
       " '值得': 1588,\n",
       " '做好': 1603,\n",
       " '应对': 4339,\n",
       " '月初': 5801,\n",
       " '为首': 881,\n",
       " '模式': 6139,\n",
       " '保持': 1527,\n",
       " '随着': 9606,\n",
       " '带来': 4218,\n",
       " '刷新': 2177,\n",
       " '高点': 9942,\n",
       " '降温': 9568,\n",
       " '加剧': 2245,\n",
       " '现象': 6936,\n",
       " '必须': 4607,\n",
       " '重视': 9381,\n",
       " '红利': 7702,\n",
       " '军工': 1896,\n",
       " '上周': 468,\n",
       " '王者': 6898,\n",
       " '荣耀': 8263,\n",
       " '类似': 7655,\n",
       " '来自': 5992,\n",
       " '比较': 6259,\n",
       " '敏感': 5395,\n",
       " '半年': 2414,\n",
       " '容易': 3894,\n",
       " '实际上': 3835,\n",
       " '美联储': 7944,\n",
       " '加息': 2253,\n",
       " '缩表': 7875,\n",
       " '行为': 8377,\n",
       " '货币政策': 8751,\n",
       " '开启': 4409,\n",
       " '流行': 6482,\n",
       " '很难': 4553,\n",
       " '中性': 821,\n",
       " '偏紧': 1600,\n",
       " '常态': 4237,\n",
       " '基本面': 3257,\n",
       " '阶段': 9522,\n",
       " '一段': 334,\n",
       " '高速': 9956,\n",
       " '增速': 3307,\n",
       " '下滑': 530,\n",
       " '而是': 7987,\n",
       " '周期': 2910,\n",
       " '适度': 9196,\n",
       " '供给': 1475,\n",
       " '结构性': 7809,\n",
       " '改革': 5342,\n",
       " '决定': 1946,\n",
       " '运行': 9062,\n",
       " '之下': 932,\n",
       " '前期': 2194,\n",
       " '单纯': 2448,\n",
       " '依托': 1477,\n",
       " '能够': 8072,\n",
       " '担忧': 5043,\n",
       " '尤其': 3998,\n",
       " '越来越': 8876,\n",
       " '甚至': 6986,\n",
       " '排除': 5168,\n",
       " '选择': 9210,\n",
       " '伴随': 1377,\n",
       " '谨慎': 8709,\n",
       " '气氛': 6296,\n",
       " '对于': 3910,\n",
       " '梳理': 6104,\n",
       " '警惕': 8536,\n",
       " '兑现': 1681,\n",
       " '利润': 2140,\n",
       " '重新': 9372,\n",
       " '机会': 5924,\n",
       " '阻力': 9524,\n",
       " '密集': 3905,\n",
       " '及时': 2587,\n",
       " '减仓': 1964,\n",
       " '毕竟': 6265,\n",
       " '通道': 9250,\n",
       " '扭转': 4938,\n",
       " '暂时': 5740,\n",
       " '持有': 5095,\n",
       " '主板': 903,\n",
       " '终于': 7769,\n",
       " '遗憾': 9283,\n",
       " '之际': 956,\n",
       " '周五': 2903,\n",
       " '怎样': 4634,\n",
       " '登记': 7131,\n",
       " '托管': 4914,\n",
       " '实施细则': 3825,\n",
       " '实施': 3823,\n",
       " '发改委': 2642,\n",
       " '抓紧': 4976,\n",
       " '制定': 2162,\n",
       " '出台': 1999,\n",
       " '价格': 1279,\n",
       " '监管': 7171,\n",
       " '规则': 8492,\n",
       " '社部': 7369,\n",
       " '月底': 5802,\n",
       " '基本': 3254,\n",
       " '实现': 3828,\n",
       " '跨省': 8909,\n",
       " '异地': 4440,\n",
       " '就医': 4004,\n",
       " '直接': 7200,\n",
       " '成品油': 4788,\n",
       " '年内': 4277,\n",
       " '下调': 536,\n",
       " '降幅': 9566,\n",
       " '国防': 3114,\n",
       " '科工': 7421,\n",
       " '混改': 6594,\n",
       " '资产': 8789,\n",
       " '证券化': 8599,\n",
       " '进程': 9139,\n",
       " '广电总局': 4327,\n",
       " '新浪': 5512,\n",
       " '微博': 4576,\n",
       " '网站': 7894,\n",
       " '关停': 1820,\n",
       " '视听': 8503,\n",
       " '节目': 8200,\n",
       " '欧美': 6172,\n",
       " '稳健': 7478,\n",
       " '国内': 3075,\n",
       " '形式': 4512,\n",
       " '整理': 5437,\n",
       " '1800': 55,\n",
       " '相比': 7222,\n",
       " '显得': 5699,\n",
       " '弱势': 4465,\n",
       " '转换': 8971,\n",
       " '随时': 9604,\n",
       " '第一': 7548,\n",
       " '第二': 7576,\n",
       " '实质性': 3832,\n",
       " '低开': 1387,\n",
       " '可能性': 2742,\n",
       " '防止': 9511,\n",
       " '极为': 6002,\n",
       " ...}"
      ]
     },
     "execution_count": 151,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vectorized.vocabulary_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {},
   "outputs": [],
   "source": [
    "import random "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [],
   "source": [
    "document_id_1, document_id_2 = random.randint(0, 1000), random.randint(0, 1000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "832"
      ]
     },
     "execution_count": 154,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "document_id_1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "713"
      ]
     },
     "execution_count": 155,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "document_id_2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'中 北京 时间 21 日 凌晨   投资者 即将 迎来 MSCI 的 2017 年度 市场 分类 评审 结果   该 机构 是否 会 将 A股 纳入 其 国际 指数 备受 外界 关注   彭博 中国 经济学家 陈世渊 认为   相较 于 把 A股 纳入 MSCI 全球 指数   作为 试点 和 折中 方案   MSCI 有 可能 先 把 A股 纳入 MSCI 中国 指数   目前 MSCI 中国 指数 主要 由 H股 等 非 A股 构成   陈世渊 在 接受 中国 证券网 记者 采访 时 表示   MSCI 认为 国内 交易所 对 海外 投资者 发行 A股 相关 产品 的 事前 审批 规定   可能 会 对 MSCI 产品 投资者 构成 投资 障碍   妨碍 目前 MSCI 指数 的 运用   而 对 国际 投资者 来说   目前 把 A股 纳入 MSCI 指数   象征意义 大于 实质 意义   他 表示   在 MSCI 的 新兴 市场 指数 和 亚洲   除去 日本   指数 里   H股 等 非 A股 已经 占到 30   的 份额   已经 具有 一定 代表性   讨论 中 的 新增 A股   在 MSCI 新兴 市场 指数 中 也 只 占 0   5   这 对 国际 投资者 寻求 更高 投资收益 或 分散 投资 风险   并 没有 特别 实质 意义   陈世渊 表示   在 MSCI 去年 所 列出 的 三个 主要 障碍 中   有 两个 已经 基本 扫清   深港 通 开通   扩大 了 市场准入   而 交易所 新 的 停牌 政策 也 限制 了 上市公司 随意性 停牌   但是 第三个 障碍   也 就是 对 MSCI 相关 产品 发行 的 事先 批准 的 规定   短期内 很难 得到 解决   目前 A股 比较稳定   没有 迫切需要 引入 外资 支持 股市   这种 情况 下   目前 政策 做出 调整 以 满足 第三个 条件 的 可能性 较 小'"
      ]
     },
     "execution_count": 156,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_content[document_id_1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'绿营 政治 人物 的 两岸 论述 近期 纷纷 转向   图片 来源   中 时 电子报   环球网 综合 报道   据 台湾   中 时 电子报   6 月 20 日 报道   针对 台南市 长赖 清德 以及 一众 绿营 政治 人物 围绕 着   九二 共识   的 表态   国民党 文传 会 副 主委 洪孟楷 20 日 上午 表示   民进党 常 说不清楚 什么 叫   九二 共识   一中 各表   但 民进党 内 近日 对 两岸 论述 却   各自 表述   起来   无论是   亲中   知中   友中   和 中   很 明显 地   民进党 内 政治 人物 的   各自 表述   他 质问   这 是否 意味 蔡 英文 已经 提前   跛脚   所以 党内 纷纷 项庄舞剑   洪孟楷 认为   蔡 英文 没有 能力 让 台湾 人民 过 好日子   又 没 办法 维持 两岸 正常 交流   所以 挑战 四起   洪孟楷 表示   邱义仁 说   模糊   是 现在 台湾 两岸 策略 最好 的 态度   还要 模糊 到 大陆 也 清楚 我们 的 模糊   这 根本就是 为 过去 一年 蔡 政府 错误 的 两岸 政策 打 模糊 仗   连 蔡 的 重要 幕僚 都 只会 模糊   那 民众 想 问   这个   政府   从上到下   到底 谁 对 两岸 政策 清楚   说不清   讲 不明   每天 只会   文青 口号 治台   就是 过去 一年 多 蔡 当局 最大 的 问题   洪 又 说   台南市 长赖 清德 也 不该 模糊 谈话   尤其 是   九二 共识   和   一国两制   两个 毫不 相连 的 名词 兜 在 一起   九二 共识   就是   九二 共识   一国两制   是   一国两制   国民党 的 主张 两岸 交流 就是   九二 共识   一中 各表   从来 没说 过   也 不 认同 两岸 问题 上要 以   一国两制   解决   从 过去 到 现在 都 不曾有过   洪孟楷 批评   赖清德 是 想争   大位   之 人   就 应该 说明 清楚   自己 的 两岸 政策 到底 为何   不要 在 台南 说 一套   在 美国 讲 一套   如果   九二 共识 不是 问题   的话   是否 愿意 当 民进党 表态 支持   九二 共识   的 第一 人   另外   他 认为   除了 赖清德 要说 清楚 自己 的 立场   蔡 英文 更 不 应该 闪躲   民进党 现在 内部 群雄 并 起 向 大陆 转向   实际上 就是 凸显 两岸关系 根本 没有 维持 蔡 英文   保证 的 现状   现在 民进党 内 提出   什么 中   路线   都 代表 着 目前 蔡 英文 在 两岸关系 上 的 处理 是 不及格 的   他 认为   不管 是   亲中   知中   和 中   就是 代表 要 正视 大陆 和 台湾 的 关系   连 民进党   立委   郭正亮 都 说 美国 调整 对 中 政策   那 台湾 又 该 装聋作哑 吗   洪孟楷 最后 称   希望 台湾 的   邦交国   不要 再断   能 继续 参与 国际 组织   两岸 稳定 和平 交流 发展   台湾 人民 安居乐业   过 好日子   请 蔡 英文 的 见识 和 高度 不要 不如 县市长   这些 过去 的 现状   请 蔡 当局 还给 人民 一年 前 的 现状   请 还给 人民 国民党 时期 的 两岸 交流 的 顺畅 现况'"
      ]
     },
     "execution_count": 157,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_content[document_id_2]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {},
   "outputs": [],
   "source": [
    "vector_of_d_1 = X[document_id_1].toarray()[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {},
   "outputs": [],
   "source": [
    "vector_of_d_2 = X[document_id_2].toarray()[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {},
   "outputs": [],
   "source": [
    "random_choose = random.randint(0, 1000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "156"
      ]
     },
     "execution_count": 161,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "random_choose"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'北京 时间 6 月 20 日 消息   日前 在 英格兰 曼彻斯特 沃斯利 高尔夫 俱乐部   Worsley   Golf   Club   举行 的 一场 男子 高尔夫 比赛 中   前 曼联 球星 瑞恩   吉格斯   Ryan   Giggs   成为 话题 人物   他 因 没交 5 英镑 的 报名费 而 被 取消 了 比赛 资格   参赛 的 球员 们 需要 在 比赛 的 前一天 交纳 5 英镑 的 现金   但 吉格斯 和 他 的 朋友 托尼   卡米 莱利   Tony   Camilleri   没 交钱 就 直接 去 比赛 了   当 比赛结果 在 沃斯利 高尔夫 俱乐部 的 网站 上 公布 时   差点 为 14 的 吉格斯 排在 144 名中 的 最后 一位   名字 前面 标着   DQ   也就是说 他 被 取消 了 比赛 资格   网站 上 还 说明 取消 他 比赛 资格 的 原因 是 他 没 支付 报名费   会所里 为此 笑声 一片   一位 比赛 者 说   吉格斯 银行 里 存款 可 不少   所以 我们 都 笑 了   他 在 俱乐部 是 很 受欢迎 的 人   但 规则 就是 规则   任何人 参加 比赛 都 要 在 俱乐部 的 专卖店 支付 5 英镑 的 现金   我 相信 这 只是 个 疏忽   但 不管怎么 说 他 还是 和 自己 的 同伴 打 了 球   看到 他 的 名字 在 比分 榜 垫底   大家 都 笑 了   我 想 他 自己 也 会 看到 这件 事情 有趣 的 一面   吉格斯 在 其 23 年 的 职业 足球 生涯 中 一直 效力 于 英超 曼联 俱乐部   并 帮助 曼联 赢得 了 13 座 英超 联赛 冠军   2 座 欧冠 冠军   4 次 英格兰 足总杯 冠军 等 36 座 冠军 奖杯   多年 来   吉格斯 一直 是 沃斯利 高尔夫 俱乐部 的 会员   会费 是 每年 950 英镑   而 在 效力 红魔 的 最后 几年   他 的 周薪 是 8 万英镑   沃斯利 高尔夫 俱乐部 随后 给 所有 的 会员 发邮件 说   我们 提醒 所有 的 会员 注意   参加 任何 比赛 都 需要 在 比赛 的 前一天 按 要求 交纳 一定 数额 的 现金   我们 得知   很多 会员 没有 带 足够 的 现金 是 很 常见 的   Colleen'"
      ]
     },
     "execution_count": 162,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_content[random_choose]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.spatial.distance import cosine"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "metadata": {},
   "outputs": [],
   "source": [
    "def distance(v1, v2): return cosine(v1, v2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0"
      ]
     },
     "execution_count": 165,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distance([1, 1], [2, 2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.994483445526207"
      ]
     },
     "execution_count": 166,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distance(X[random_choose].toarray()[0], X[document_id_1].toarray()[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9757075222629863"
      ]
     },
     "execution_count": 167,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "distance(X[random_choose].toarray()[0], X[document_id_2].toarray()[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'点击 图片 进入 下 一页   庭审 现场   李小华   摄   中新网 自贡 6 月 23 日电   刘刚   李小华   叶 卫东   23 日 下午   四川 自贡市 中级 人民法院 集中 公开 宣判 了 4 起 毒品 案件   毒品 总量 约 为 15890 克   涉案 被告人 16 人   其中   判处 无期徒刑 2 人   十年 以上 有期徒刑 5 人   据悉   目前   自贡市 制造 毒品 案件 增长 较 快且 数量 大   仅 今年 上半年   就 破获 3 件 制造 毒品 数量 巨大 的 案件   自贡市 中院 新闻 发言人 在 随后 召开 的 新闻 通气会 上 披露   2016 年 6 月 到 2017 年 5 月   全市 两级 法院 共 审理 毒品 犯罪案件 213 件   涉案 被告人 250 人   同比 上升 11   和 14   其中   贩卖   运输 案件 117 件   涉案 被告人 132 人   制造 毒品 案件 18 件   涉案 被告人 31 人   非法 持有 毒品   容留 他人 吸毒 案件 78 件   涉案 被告人 89 人   点击 图片 进入 下 一页   据介绍   自贡市 毒品 主要 来源于 国外   国内 两条 线路   国外 主要 是 老挝   缅甸   国内 主要 是 广东   云南   此外   近年来 出现 的 新型 毒品 也 造成 了 极大 的 社会 危害'"
      ]
     },
     "execution_count": 168,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_content[320]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'央广网 西安 6 月 19 日 消息   记者 雷恺   陕西省 今年 将 招录 250 名 中央财政 支持 中西部 农村 订单 定向 免费 本科 医学生   其中 临床医学 200 名   中医学 50 名   培养 计划 重点 向 贫困地区   革命 老区 倾斜   据介绍   国家 免费 医学生 只 招收 农村 学生   报考者 须 同时 具备 两个 条件   一是 已 参加 2017 年 陕西省 普通高校 招生 考试 报名   二是 本人 及 父亲 或 母亲 或 法定 监护人 户籍地 须 在 陕西省 农村   本人 具有 当地 连续 3 年 以上 户籍   国家 免费 医学生 录取 后   获得 入学 通知书 前   须 与 培养 高校 和 定向 就业 所在地 的 县级 卫生 计生   人力资源 社会保障 行政部门 签署 统一 制式 的 定向培养 和 就业 协议   承诺 毕业 后 到 有关 基层 医疗卫生 机构 服务 6 年   含 3 年 全科 专业 住院医师 规范化 培训   未 按 入学 前 签署 的 协议 规定 履约 的   退还 已 减免 的 教育 培训 费用   生活 补助 和 以上 费用 50   的 违约金   履约 情况 纳入 医师 诚信 管理   公布 违约 记录   并 记入 人事档案   国家 免费 医学生 毕业 后 拒绝 履约 的   5 年内 不得 应聘 全省 行政 范围 内 的 各级 各类 事业单位   免责 声明   本文 仅 代表 作者 个人观点   与 环球网 无关   其 原创性 以及 文中 陈述 文字 和 内容 未经 本站 证实   对 本文 以及 其中 全部 或者 部分 内容   文字 的 真实性   完整性   及时性 本站 不 作 任何 保证 或 承诺   请 读者 仅作 参考   并 请 自行 核实 相关 内容'"
      ]
     },
     "execution_count": 169,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_content[72]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'央广网 烟台 6 月 19 日 消息   记者 刘颖超   通讯员 蒋毅   牟 辰龙   6 月 17 日 下午   山东 烟台 系山 边防 派出所 民警 来到 荣成市 寻 山镇   将 涉嫌 非法拘禁 的 犯罪 嫌疑人 曲某 抓获   至此   4 名 涉嫌 非法拘禁 的 犯罪 嫌疑人 已 全部 落网   据 了解   家住 龙湖 葡 醍 海湾 小区 的 夏 某于 6 月 15 日 上午 带 儿子 外出 期间   在 其 地下 车库 被 四名 从 威海 荣成 赶来 要账 的 鞠 某 等 四人 截住   鞠某 以夏 某 拖欠 工程款 为 由   要求 与 夏 某 找个 地方 好好 谈谈   但 夏 某 不 认为 自己 拖欠 鞠某 等 人 工程款   且 有 法院 判决   遂 拒 不 配合   鞠某 等 人 情绪 激烈   强行 将 夏 某 拖拽 到 车上   并 对 夏 某 进行 殴打   要求 夏某 配合 他们 找 地方 谈 拖欠 工程款 的 问题   鞠某 等 人 驱车 强行 将 夏 某 带到 了 威海 经济 技术开发区   但 看到 躺 在 车 后排 的 夏 某 一直 处于 迷糊 状态   心中 也 产生 了 担忧   于是 又 驱车 从 威海 将 鞠 某 送回 了 烟台市 牟平区 龙湖 葡 醍 海湾 小区   此时   山东 烟台 系山 边防 派出所 民警 正 对照 监控 逐个 排查 外地 可疑 车辆   看到 鞠某 等 人 驱车 返回   遂 传唤 到 派出所 接受 调查   经 询问   曲某 在 乘车 到达 威海 时 就 已 下车   而 鞠 某 等 人 以为 这次 来 牟平 只是 单纯 要账   没想到 自己 的 行为 却 触犯 了 刑法 有关 规定   他们 对 自己 不 懂法 的 行为 后悔不已   目前   鞠某 等 人 因涉嫌 非法拘禁 罪 被 公安机关 依法 刑事拘留   免责 声明   本文 仅 代表 作者 个人观点   与 环球网 无关   其 原创性 以及 文中 陈述 文字 和 内容 未经 本站 证实   对 本文 以及 其中 全部 或者 部分 内容   文字 的 真实性   完整性   及时性 本站 不 作 任何 保证 或 承诺   请 读者 仅作 参考   并 请 自行 核实 相关 内容'"
      ]
     },
     "execution_count": 170,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_content[85]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'原 标题   武汉 警方 一下子 抓 了 808 人   还 都 是 俊男靓女   原来 他们 每天 偷偷摸摸 干 这事   武汉 警方 一下子 抓 了 808 人   还 都 是 俊男靓女   原来 他们 每天 偷偷摸摸 干 这事   大家 还 记得 今年 4 月 光谷 曾 发生 的 一件 大事 吗   当时 微博 和 朋友圈 上 都 传 疯 了   说 是 光谷 国际 广场 大量 白领 被捕   现场 手铐 都 不够 用   结果 经 记者 深入调查 后 发现   这些 被 抓 的 并 不是 网传 的   白领   而是 一个 有 282 名 涉案人员 的 诈骗 团伙   天呐   接近 300 人 的 犯罪团伙   你 以为 这个 数字 已经 够 庞大 了 吗   错   还有 让 你 更 震惊 的   据 武汉市公安局 通报   6 月 9 日   警方 在 武汉三镇 的 10 处 窝点 内   抓获 了 808 名 嫌疑人   而且 令人 大跌眼镜 的 是   他们 当中 大多 都 是 年青人   用   俊男靓女   来 形容 也 不为过   年纪 不 大   外形 不差   那么 到底 是 犯 了 什么 事 呢   其实 说 起来   这 事 估计 绝大多数 人 都 碰到 过   那 就是 网络 诈骗   涉案 电脑 800 余台   手机 2000 余部   银行卡 3000 余张   因为 涉及 的 账户 太 多   涉案 金额 目前 还 没 统计 出来   这是 全国 涉案人 数最多 的 一起 特大 网络 诈骗案   这些 惊人 的 数字 背后   究竟 藏 着 什么样 的 秘密   警方 把 他们 一锅端 了   今年 4 月   武汉 警方 获悉 在 塔子 湖 地区 有 电信 网络 诈骗 活动   经过 2 个 多月 的 秘密 侦查   发现 了 一个 特大 网络 诈骗 犯罪团伙   原来 该 团伙 利用 非法 互联网 金融 平台   实施 推荐 股票   炒 期货 等 诈骗   其 10 处 窝点 散布 在 江岸   洪山   东湖 高新   硚口   江汉 等 5 个区 的 写字楼 内   涉案人员 多达 800 余名   6 月 9 日 上午   武汉 警方 组织 了 1000 余名 警力   分成 14 个 抓捕 组   同时 对 首脑 分子 和 10 处 窝点 实施 了 突击 抓捕 行动   目前   警方 已 依法 刑事拘留 411 人   对 45 人 取保候审   该案 还 在 进一步 调查 中   他们 是 如何 进行 诈骗 的   人数 众多 的 这个 诈骗 组织 是 怎么样 进行 人员 分工   又 是 如何 一步步 引 受害人 上钩 的 呢   2015 年 以来   该 犯罪团伙 以 投资 组建 的 某 集团公司 为 总部   旗下 成立 多家 分公司   子公司   设立 人事部   财务部   行政部 等 机构 部门   分为 经理   分析师   业务 小组长   业务员 等 组织 层级   在 网上 发布 招聘 信息 招揽   培训 业务员   依托 虚构 的 互联网 金融 平台   通过 业务员 散布   有 内幕消息   等 虚假 信息   引诱 受害人 加入 专门 的 QQ 群   微信群   再 由 所谓 的   证券公司 资深 分析师   和   经理   逐个 联系 受害人   引诱 受害人 交易 股票   期货   贵金属   外汇 等 金融 产品   受害人 若 获利 则 按   五五   六四   分成   若 亏损 则 拉入   黑名单   不再 联系   网友 热评   李晓猫 不吃 鱼   在 人 信汇 写字楼 面试 过 一个 这种 工作   一听 不对劲 就 撤 了   说 不 适合 我   我 嘴巴 笨   习惯 就 好 33xxzz   八百多 人   可能 相当 部分 人 都 不大清楚 自己 到底 是 啥 性质 工作 吧   刑法 守望者   要 警惕 了   普法   各种 工作 都 要 预防 法律 风险   要 做 个 明白人   不是 什么 钱 都 可以 拿   尤其 是 从事 金融   保险   银行   财务 的 人员   不 小心 就 触犯 刑法   童颜 永 在   心理医生   盘 在 写字楼 内 的 诈骗 公司   以 高薪 招聘 业务人员 为 诱饵   实行 诈骗 的 顽疾   终于 被 揣 掉 了   你 的 渔船   联系 一下 自己 的 朋友   看 有 多少 个 被 抓 了   The   former   degrees   招聘 平台 和 销售 会上 百分之八十 就是 这种 工作   从 源头 抓起   防止 更多人 受伤   宇宇家 的 小 仙女   恐怖   进 了 一个 星期 发现 不太 对劲   这 尼玛 是 让 我 骗人 啊   赶紧 自己 跑 路 了   z _ 路 漫漫   我 做 过   后来 听说 公司 跑 路 了   睡 不 饱 怎么办 啊 小铭大   我 都 面试 过 这种 工作   还 培训 了 几天   话术 一看 就 的 套路 骗   果断 闪 人   这种 公司 太多 了   遗忘 的 金 叶子   碰到 这样 的 骗子 找 我们 做 视频   最后 被 我 打电话 骂 了 一顿   a 一元复始 c   还有 那些 天天 电话 骚扰 别人 信贷 的   卖房 的   卖 邮票 的   发票 的 社会 生活 中 一不留神 就 会 被 坑   所以 活着 需要 理智   笑 看 世间 丑 2448337651   这 是 在 路上 发骗 广告 搞 高利贷 的 那些 互联网 金融公司 吗   未来 温暖 的 时光 属于 自己   金融 港   街道口 都 是 这些   好多 同学 刚刚 毕业 都 去 做 过   都 是 网上 投 简历   这些 公司 就 猛 联系 你   其实 好多 都 不 知道 这是 骗人 的 吧   有 强迫症 的 天蝎座 怪 咖   吓 到 我 了   冉冉 的 彗星 是 个 宝   早上 还 接到 说 做 期货   霖心 金鱼 And 张蒙   都 不敢 去 找 工作 了   蒋 妈妈 1   5   刚回到 武汉   准备 找份 工作   看 网上 好多 这种 招聘 的   感觉 像 骗人 的   都 不 看 的   看 了 网友 的 评论   小编 只有 一个 感觉   找 工作 时 一定 要 擦亮 双眼   不要 因为 一时 疏忽   而 给 自己 的 生活 抹 上 污点   俗话说   知己知彼 百战不殆   弄清 网络 诈骗 的 常见 手法   才能 避免 上当受骗   下面 就 为 大家 奉上 防骗 攻略   遇到 诈骗 只 需 见招拆招   网络 防骗 攻略   电信 网络 诈骗 犯罪 往往 表现 为   六大 陷阱   1   物美价廉 的 网购 陷阱   2   推销产品 的 便利 陷阱   3   异性 交友 的 情感 陷阱   4   中奖 信息 的 惊喜 陷阱   5   冒充 国家 工作人员 的 恐吓 陷阱   6   钓鱼 网站 和 网上银行 的 迷惑 陷阱   下面 是 网络 诈骗 中 比较 常见 的 几种 类型   一起 学习 如何 防范   让 骗子 无处 遁形   1 虚假 红包 诈骗   临近 节日 时   各大 电商 平台 会以 派 发红包 的 形式   为 促销 活动 预热   有 不法分子 趁机 在 微信 朋友圈   微信 群 等 平台 中   派发 虚假 红包   诱骗 大家 点击   套取 个人信息   或 传播 木马病毒   盗取 银行账号   识骗 方法   慎点 此类 红包 领取 链接   如果 不 小心 点击   应 第一 时间 关闭 手机 网络   修改 网银   支付宝 等 重要 账户 密码   并 通过 安全软件 查杀 木马病毒   2 二维码 诈骗   抢购 高峰期   网页 打开 慢   网银 交易 出现 比较 迟钝 现象   有 不法分子 会 冒充 客服 或 店主   以 单子 出现   调单   卡单   等 说法 为 幌子   故意 说 没有 收到 款项 欺诈 消费者   取得 信任 后 要求 消费者 以 其他 办法 进行 转账   识骗 方法   消费者 要 注意   调单   卡单   激活 订单 等 多 为 诈骗 术语   见 此类 词语 基本 可以 断定 对方 是 骗子   没有 收到 款项 这样 的 问题   是 不会 在 东京   天猫 这样 的 大型 网络 购物 平台 出现 的   3 钓鱼 网站 诈骗   节假日 期间   公众 的 购物 热情高涨   假冒 正规 电商 官方网站   在线 支付 平台 等 的 钓鱼 网站 活动 频繁   消费者 在 忙于 浏览 各色 商品 页面 时   往往 容易 忽略 链接 的 安全性   从而 误入 诈骗 分子 发来 的 虚假 钓鱼 网站   识骗 方法   对 陌生人 发来 的 商品 链接 要 提高警惕   谨慎 打开   如遇 网络 诈骗 应 第一 时间 向 属地 公安机关 报案   并 向 举报中心 举报   4 虚假 客服 诈骗   抢购 高峰期   网页 打开 慢   网银 交易 出现 比较 迟钝 现象   有 不法分子 会 冒充 客服 或 店主   以 单子 出现   调单   卡单   等 说法 为 幌子   故意 说 没有 收到 款项 欺诈 消费者   取得 信任 后 要求 消费者 以 其他 办法 进行 转账   识骗 方法   消费者 要 注意   调单   卡单   激活 订单 等 多 为 诈骗 术语   见 此类 词语 基本 可以 断定 对方 是 骗子   没有 收到 款项 这样 的 问题   是 不会 在 东京   天猫 这样 的 大型 网络 购物 平台 出现 的   5 预购 诈骗   还有 不法分子 会以   提前 购   预售   为 诱饵 向 用户 发送 木马 链接   除了 网购 链接 被 植入 木马   钓鱼 网站   还有 不法分子 会 在   预售   活动 上 做文章   骗取 买家   定金   识骗 方法   慎点 不明 链接   有关 购物 短信 也 不要 轻信   同时 也 请 给 手机 装 上 杀毒软件   提供 安全 的 上网 环境   6 中奖 诈骗   购物 过后   很多 人会 陆陆续续 收到 一些 商家 的 中奖 信息   切记 不要 随意 拨打 中奖 信息 提供 的 电话号码 或 点击 发来 的 网站 链接 地址   很 可能 会 让 手机 或 电脑 感染 上 木马病毒   识骗 方法   收到 中奖 信息 后 留心 一下   看一看 你 所 购买 的 产品 中 有没有 这家 的 产品   然后 最好 能 在 天猫 店里 直接 和 卖家 联系 确认 消息   7 朋友圈 诈骗   由于 缺少 第三方 监管 平台   不法分子 会 在 朋友圈 发布 虚假 低价 购物 信息   骗取 消费者 汇款 购买   识骗 方法   朋友 圈里 的 销售 行为   存在 取证难   查处 难 的 问题   因此 建议 大家 在 朋友圈 买 东西 时   尽可能 选择 正规 第三方 支付 平台 进行 支付   避免 直接 支付   以 确保安全   8 快递 诈骗   不法分子 会 从 非法 渠道 获得 客户 信息   制作 假 的 快递 单   随后 冒充 快递 员 提前 联系 受骗者   把 虚假 货品 送到 消费者 手中   骗取钱财   识骗 方法   签收 包裹 前   务必 按 官方 的 信息 核实 对方 身份   并 确认 是否是 自己 的 包裹   在 取货 之后   一定 记得 把 自己 的 个人信息 从 包裹 上 完全 抹 去   避免 个人信息 泄露   9 退款 诈骗   不法分子 常 以   退款   或   退货   为 由 假扮 网店 客服   要求 消费者 点击 其 提供 的   钓鱼 网站   链接   盗取 持卡人 的 银行 密码 及 动态 验证码   识骗 方法   如果 接到 网上 购物 退款 的 电话 或 短信 等 信息   一定 要 谨慎 对待   切莫 泄露 银行卡 账号   户名   动态 验证码 等 个人信息   同时   提醒 大家   保护 好 自己 账户 信息   遇到 情况 一定 及时 报警   武汉 破 特大 网络 诈骗案 抓 808 人   以   内幕消息   骗人   6 月 12 日   武汉市公安局 向 澎湃 新闻 通报   该局 破获 一起 特大 网络 诈骗案   抓获 包括 卢某   熊某 等 8 名 首脑 分子 在内 的 涉案人员 808 名   查扣 涉案 电脑 800 余台   手机 2000 余部   银行卡 3000 余张   据悉   这是 全国 打击 利用 非法 互联网 金融 平台 实施 诈骗 犯罪 中   查获 涉案人 数最多 的 一起 案件'"
      ]
     },
     "execution_count": 171,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "news_content[8]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\wanga26\\AppData\\Local\\Continuum\\anaconda3\\envs\\cs231n\\lib\\site-packages\\scipy\\spatial\\distance.py:698: RuntimeWarning: invalid value encountered in double_scalars\n",
      "  dist = 1.0 - uv / np.sqrt(uu * vv)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[352,\n",
       " 1126,\n",
       " 301,\n",
       " 59,\n",
       " 69,\n",
       " 28,\n",
       " 35,\n",
       " 33,\n",
       " 99,\n",
       " 48,\n",
       " 81,\n",
       " 109,\n",
       " 43,\n",
       " 42,\n",
       " 92,\n",
       " 108,\n",
       " 38,\n",
       " 49,\n",
       " 8,\n",
       " 24,\n",
       " 93,\n",
       " 116,\n",
       " 80,\n",
       " 77,\n",
       " 37,\n",
       " 52,\n",
       " 85,\n",
       " 86,\n",
       " 36,\n",
       " 22,\n",
       " 110,\n",
       " 13,\n",
       " 34,\n",
       " 65,\n",
       " 83,\n",
       " 7,\n",
       " 64,\n",
       " 98,\n",
       " 105,\n",
       " 11,\n",
       " 73,\n",
       " 54,\n",
       " 61,\n",
       " 9,\n",
       " 21,\n",
       " 27,\n",
       " 4,\n",
       " 91,\n",
       " 95,\n",
       " 20,\n",
       " 46,\n",
       " 101,\n",
       " 12,\n",
       " 0,\n",
       " 119,\n",
       " 6,\n",
       " 84,\n",
       " 115,\n",
       " 97,\n",
       " 29,\n",
       " 70,\n",
       " 71,\n",
       " 45,\n",
       " 113,\n",
       " 51,\n",
       " 62,\n",
       " 96,\n",
       " 30,\n",
       " 14,\n",
       " 26,\n",
       " 41,\n",
       " 56,\n",
       " 72,\n",
       " 66,\n",
       " 58,\n",
       " 25,\n",
       " 67,\n",
       " 44,\n",
       " 31,\n",
       " 55,\n",
       " 5,\n",
       " 106,\n",
       " 53,\n",
       " 94,\n",
       " 102,\n",
       " 47,\n",
       " 90,\n",
       " 40,\n",
       " 114,\n",
       " 60,\n",
       " 16,\n",
       " 112,\n",
       " 10,\n",
       " 75,\n",
       " 88,\n",
       " 74,\n",
       " 111,\n",
       " 15,\n",
       " 1,\n",
       " 78,\n",
       " 23,\n",
       " 57,\n",
       " 17,\n",
       " 19,\n",
       " 79,\n",
       " 63,\n",
       " 82,\n",
       " 68,\n",
       " 50,\n",
       " 104,\n",
       " 117,\n",
       " 89,\n",
       " 118,\n",
       " 2,\n",
       " 3,\n",
       " 18,\n",
       " 32,\n",
       " 39,\n",
       " 87,\n",
       " 100,\n",
       " 103,\n",
       " 156,\n",
       " 466,\n",
       " 155,\n",
       " 278,\n",
       " 921,\n",
       " 529,\n",
       " 506,\n",
       " 341,\n",
       " 223,\n",
       " 1073,\n",
       " 1049,\n",
       " 1061,\n",
       " 1110,\n",
       " 1109,\n",
       " 1101,\n",
       " 1116,\n",
       " 967,\n",
       " 1069,\n",
       " 984,\n",
       " 1075,\n",
       " 1076,\n",
       " 977,\n",
       " 1063,\n",
       " 973,\n",
       " 965,\n",
       " 1112,\n",
       " 981,\n",
       " 1107,\n",
       " 1056,\n",
       " 1095,\n",
       " 1053,\n",
       " 975,\n",
       " 1119,\n",
       " 1043,\n",
       " 1103,\n",
       " 971,\n",
       " 1086,\n",
       " 1065,\n",
       " 996,\n",
       " 1111,\n",
       " 1087,\n",
       " 1118,\n",
       " 987,\n",
       " 972,\n",
       " 1067,\n",
       " 1093,\n",
       " 998,\n",
       " 1089,\n",
       " 979,\n",
       " 1044,\n",
       " 1092,\n",
       " 1040,\n",
       " 1080,\n",
       " 1070,\n",
       " 989,\n",
       " 1106,\n",
       " 1050,\n",
       " 1094,\n",
       " 1097,\n",
       " 1060,\n",
       " 1079,\n",
       " 986,\n",
       " 968,\n",
       " 994,\n",
       " 1088,\n",
       " 1046,\n",
       " 985,\n",
       " 1062,\n",
       " 1081,\n",
       " 1082,\n",
       " 988,\n",
       " 1078,\n",
       " 963,\n",
       " 1117,\n",
       " 993,\n",
       " 961,\n",
       " 962,\n",
       " 992,\n",
       " 1059,\n",
       " 1113,\n",
       " 1098,\n",
       " 1114,\n",
       " 1051,\n",
       " 1077,\n",
       " 969,\n",
       " 1047,\n",
       " 976,\n",
       " 978,\n",
       " 1105,\n",
       " 1074,\n",
       " 983,\n",
       " 1052,\n",
       " 1102,\n",
       " 1104,\n",
       " 1045,\n",
       " 1041,\n",
       " 1048,\n",
       " 1115,\n",
       " 960,\n",
       " 1042,\n",
       " 1096,\n",
       " 1054,\n",
       " 1057,\n",
       " 1066,\n",
       " 991,\n",
       " 1068,\n",
       " 1072,\n",
       " 999,\n",
       " 970,\n",
       " 982,\n",
       " 1090,\n",
       " 1055,\n",
       " 1099,\n",
       " 1064,\n",
       " 1058,\n",
       " 1085,\n",
       " 1108,\n",
       " 1083,\n",
       " 974,\n",
       " 1071,\n",
       " 980,\n",
       " 1091,\n",
       " 1084,\n",
       " 1100,\n",
       " 964,\n",
       " 966,\n",
       " 990,\n",
       " 995,\n",
       " 997,\n",
       " 1272,\n",
       " 229,\n",
       " 418,\n",
       " 694,\n",
       " 472,\n",
       " 484,\n",
       " 205,\n",
       " 499,\n",
       " 574,\n",
       " 805,\n",
       " 535,\n",
       " 162,\n",
       " 509,\n",
       " 481,\n",
       " 377,\n",
       " 558,\n",
       " 657,\n",
       " 507,\n",
       " 735,\n",
       " 387,\n",
       " 369,\n",
       " 191,\n",
       " 144,\n",
       " 354,\n",
       " 147,\n",
       " 676,\n",
       " 390,\n",
       " 133,\n",
       " 441,\n",
       " 510,\n",
       " 738,\n",
       " 316,\n",
       " 449,\n",
       " 339,\n",
       " 299,\n",
       " 312,\n",
       " 253,\n",
       " 261,\n",
       " 688,\n",
       " 432,\n",
       " 1215,\n",
       " 175,\n",
       " 548,\n",
       " 123,\n",
       " 444,\n",
       " 161,\n",
       " 732,\n",
       " 210,\n",
       " 384,\n",
       " 518,\n",
       " 410,\n",
       " 198,\n",
       " 616,\n",
       " 686,\n",
       " 488,\n",
       " 895,\n",
       " 517,\n",
       " 827,\n",
       " 673,\n",
       " 1256,\n",
       " 520,\n",
       " 292,\n",
       " 180,\n",
       " 754,\n",
       " 197,\n",
       " 1200,\n",
       " 255,\n",
       " 538,\n",
       " 269,\n",
       " 527,\n",
       " 545,\n",
       " 171,\n",
       " 909,\n",
       " 379,\n",
       " 371,\n",
       " 297,\n",
       " 647,\n",
       " 798,\n",
       " 711,\n",
       " 534,\n",
       " 880,\n",
       " 513,\n",
       " 700,\n",
       " 483,\n",
       " 1020,\n",
       " 708,\n",
       " 430,\n",
       " 415,\n",
       " 228,\n",
       " 737,\n",
       " 211,\n",
       " 564,\n",
       " 536,\n",
       " 950,\n",
       " 221,\n",
       " 192,\n",
       " 886,\n",
       " 467,\n",
       " 834,\n",
       " 774,\n",
       " 531,\n",
       " 447,\n",
       " 938,\n",
       " 782,\n",
       " 425,\n",
       " 914,\n",
       " 124,\n",
       " 521,\n",
       " 942,\n",
       " 1036,\n",
       " 363,\n",
       " 421,\n",
       " 911,\n",
       " 1224,\n",
       " 1158,\n",
       " 300,\n",
       " 665,\n",
       " 208,\n",
       " 251,\n",
       " 128,\n",
       " 323,\n",
       " 142,\n",
       " 199,\n",
       " 892,\n",
       " 1026,\n",
       " 1192,\n",
       " 1228,\n",
       " 398,\n",
       " 347,\n",
       " 1164,\n",
       " 254,\n",
       " 1203,\n",
       " 195,\n",
       " 1173,\n",
       " 780,\n",
       " 885,\n",
       " 1260,\n",
       " 859,\n",
       " 237,\n",
       " 356,\n",
       " 243,\n",
       " 396,\n",
       " 494,\n",
       " 163,\n",
       " 751,\n",
       " 404,\n",
       " 454,\n",
       " 925,\n",
       " 1132,\n",
       " 475,\n",
       " 525,\n",
       " 468,\n",
       " 1168,\n",
       " 959,\n",
       " 310,\n",
       " 800,\n",
       " 170,\n",
       " 1212,\n",
       " 828,\n",
       " 196,\n",
       " 578,\n",
       " 459,\n",
       " 795,\n",
       " 127,\n",
       " 899,\n",
       " 364,\n",
       " 554,\n",
       " 436,\n",
       " 406,\n",
       " 407,\n",
       " 426,\n",
       " 786,\n",
       " 530,\n",
       " 452,\n",
       " 503,\n",
       " 1139,\n",
       " 533,\n",
       " 289,\n",
       " 231,\n",
       " 1121,\n",
       " 504,\n",
       " 919,\n",
       " 883,\n",
       " 821,\n",
       " 522,\n",
       " 1022,\n",
       " 1270,\n",
       " 478,\n",
       " 209,\n",
       " 759,\n",
       " 458,\n",
       " 1253,\n",
       " 725,\n",
       " 1258,\n",
       " 500,\n",
       " 139,\n",
       " 693,\n",
       " 872,\n",
       " 851,\n",
       " 675,\n",
       " 773,\n",
       " 1250,\n",
       " 609,\n",
       " 638,\n",
       " 691,\n",
       " 214,\n",
       " 1133,\n",
       " 789,\n",
       " 1241,\n",
       " 391,\n",
       " 486,\n",
       " 863,\n",
       " 349,\n",
       " 246,\n",
       " 822,\n",
       " 881,\n",
       " 1217,\n",
       " 1186,\n",
       " 833,\n",
       " 546,\n",
       " 603,\n",
       " 619,\n",
       " 372,\n",
       " 793,\n",
       " 277,\n",
       " 491,\n",
       " 502,\n",
       " 561,\n",
       " 143,\n",
       " 592,\n",
       " 505,\n",
       " 138,\n",
       " 896,\n",
       " 172,\n",
       " 778,\n",
       " 129,\n",
       " 218,\n",
       " 233,\n",
       " 560,\n",
       " 661,\n",
       " 1243,\n",
       " 381,\n",
       " 1161,\n",
       " 399,\n",
       " 259,\n",
       " 1257,\n",
       " 492,\n",
       " 416,\n",
       " 464,\n",
       " 1237,\n",
       " 234,\n",
       " 656,\n",
       " 1225,\n",
       " 1214,\n",
       " 286,\n",
       " 350,\n",
       " 660,\n",
       " 189,\n",
       " 1252,\n",
       " 498,\n",
       " 692,\n",
       " 867,\n",
       " 283,\n",
       " 333,\n",
       " 380,\n",
       " 634,\n",
       " 1218,\n",
       " 1030,\n",
       " 690,\n",
       " 188,\n",
       " 792,\n",
       " 140,\n",
       " 385,\n",
       " 713,\n",
       " 903,\n",
       " 176,\n",
       " 1025,\n",
       " 495,\n",
       " 843,\n",
       " 382,\n",
       " 346,\n",
       " 748,\n",
       " 1013,\n",
       " 366,\n",
       " 540,\n",
       " 248,\n",
       " 479,\n",
       " 241,\n",
       " 461,\n",
       " 1222,\n",
       " 1206,\n",
       " 1267,\n",
       " 405,\n",
       " 497,\n",
       " 157,\n",
       " 456,\n",
       " 414,\n",
       " 526,\n",
       " 1254,\n",
       " 485,\n",
       " 642,\n",
       " 736,\n",
       " 1259,\n",
       " 818,\n",
       " 1159,\n",
       " 740,\n",
       " 824,\n",
       " 348,\n",
       " 336,\n",
       " 190,\n",
       " 267,\n",
       " 177,\n",
       " 875,\n",
       " 703,\n",
       " 946,\n",
       " 696,\n",
       " 945,\n",
       " 443,\n",
       " 216,\n",
       " 220,\n",
       " 487,\n",
       " 669,\n",
       " 1035,\n",
       " 935,\n",
       " 1211,\n",
       " 858,\n",
       " 463,\n",
       " 275,\n",
       " 905,\n",
       " 1125,\n",
       " 1239,\n",
       " 438,\n",
       " 1123,\n",
       " 470,\n",
       " 462,\n",
       " 890,\n",
       " 337,\n",
       " 930,\n",
       " 849,\n",
       " 166,\n",
       " 910,\n",
       " 1245,\n",
       " 617,\n",
       " 537,\n",
       " 482,\n",
       " 1207,\n",
       " 342,\n",
       " 238,\n",
       " 745,\n",
       " 709,\n",
       " 760,\n",
       " 429,\n",
       " 622,\n",
       " 200,\n",
       " 1183,\n",
       " 245,\n",
       " 451,\n",
       " 837,\n",
       " 624,\n",
       " 860,\n",
       " 706,\n",
       " 590,\n",
       " 777,\n",
       " 351,\n",
       " 145,\n",
       " 855,\n",
       " 842,\n",
       " 1182,\n",
       " 742,\n",
       " 776,\n",
       " 493,\n",
       " 186,\n",
       " 841,\n",
       " 431,\n",
       " 450,\n",
       " 1198,\n",
       " 389,\n",
       " 704,\n",
       " 1027,\n",
       " 358,\n",
       " 1242,\n",
       " 408,\n",
       " 411,\n",
       " 423,\n",
       " 460,\n",
       " 1196,\n",
       " 409,\n",
       " 428,\n",
       " 455,\n",
       " 953,\n",
       " 836,\n",
       " 618,\n",
       " 514,\n",
       " 614,\n",
       " 575,\n",
       " 844,\n",
       " 768,\n",
       " 1148,\n",
       " 1264,\n",
       " 1263,\n",
       " 134,\n",
       " 225,\n",
       " 512,\n",
       " 775,\n",
       " 508,\n",
       " 469,\n",
       " 1194,\n",
       " 1032,\n",
       " 927,\n",
       " 318,\n",
       " 809,\n",
       " 185,\n",
       " 915,\n",
       " 422,\n",
       " 907,\n",
       " 473,\n",
       " 1016,\n",
       " 528,\n",
       " 752,\n",
       " 1165,\n",
       " 173,\n",
       " 135,\n",
       " 1227,\n",
       " 1266,\n",
       " 755,\n",
       " 266,\n",
       " 413,\n",
       " 702,\n",
       " 207,\n",
       " 424,\n",
       " 121,\n",
       " 1202,\n",
       " 1262,\n",
       " 784,\n",
       " 271,\n",
       " 1174,\n",
       " 1140,\n",
       " 726,\n",
       " 864,\n",
       " 490,\n",
       " 632,\n",
       " 374,\n",
       " 1240,\n",
       " 523,\n",
       " 650,\n",
       " 611,\n",
       " 697,\n",
       " 489,\n",
       " 306,\n",
       " 361,\n",
       " 906,\n",
       " 402,\n",
       " 294,\n",
       " 1220,\n",
       " 252,\n",
       " 362,\n",
       " 1028,\n",
       " 178,\n",
       " 1135,\n",
       " 359,\n",
       " 722,\n",
       " 367,\n",
       " 952,\n",
       " 1160,\n",
       " 1137,\n",
       " 164,\n",
       " 916,\n",
       " 417,\n",
       " 419,\n",
       " 956,\n",
       " 948,\n",
       " 1247,\n",
       " 230,\n",
       " 939,\n",
       " 626,\n",
       " 1232,\n",
       " 901,\n",
       " 772,\n",
       " 749,\n",
       " 1017,\n",
       " 1246,\n",
       " 643,\n",
       " 1167,\n",
       " 566,\n",
       " 879,\n",
       " 1279,\n",
       " 544,\n",
       " 714,\n",
       " 1038,\n",
       " 636,\n",
       " 850,\n",
       " 137,\n",
       " 1142,\n",
       " 148,\n",
       " 1216,\n",
       " 1172,\n",
       " 206,\n",
       " 1205,\n",
       " 1007,\n",
       " 343,\n",
       " 227,\n",
       " 807,\n",
       " 393,\n",
       " 1255,\n",
       " 541,\n",
       " 615,\n",
       " 936,\n",
       " 1002,\n",
       " 236,\n",
       " 1122,\n",
       " 1021,\n",
       " 716,\n",
       " 328,\n",
       " 181,\n",
       " 651,\n",
       " 929,\n",
       " 288,\n",
       " 723,\n",
       " 501,\n",
       " 699,\n",
       " 924,\n",
       " 1179,\n",
       " 1006,\n",
       " 826,\n",
       " 628,\n",
       " 668,\n",
       " 739,\n",
       " 1197,\n",
       " 812,\n",
       " 920,\n",
       " 496,\n",
       " 247,\n",
       " 683,\n",
       " 543,\n",
       " 621,\n",
       " 1181,\n",
       " 302,\n",
       " 1147,\n",
       " 635,\n",
       " 1153,\n",
       " 239,\n",
       " 434,\n",
       " 515,\n",
       " 644,\n",
       " 868,\n",
       " 344,\n",
       " 862,\n",
       " 857,\n",
       " 322,\n",
       " 360,\n",
       " 685,\n",
       " 1177,\n",
       " 226,\n",
       " 913,\n",
       " 465,\n",
       " 303,\n",
       " 802,\n",
       " 555,\n",
       " 375,\n",
       " 159,\n",
       " 222,\n",
       " 345,\n",
       " 1238,\n",
       " 201,\n",
       " 847,\n",
       " 242,\n",
       " 457,\n",
       " 183,\n",
       " 307,\n",
       " 476,\n",
       " 388,\n",
       " 766,\n",
       " 524,\n",
       " 918,\n",
       " 1129,\n",
       " 412,\n",
       " 174,\n",
       " 1031,\n",
       " 1195,\n",
       " 446,\n",
       " 1120,\n",
       " 670,\n",
       " 1163,\n",
       " 1187,\n",
       " 179,\n",
       " 125,\n",
       " 878,\n",
       " 439,\n",
       " 730,\n",
       " 771,\n",
       " 680,\n",
       " 1003,\n",
       " 698,\n",
       " 762,\n",
       " 1033,\n",
       " 876,\n",
       " 295,\n",
       " 602,\n",
       " 153,\n",
       " 146,\n",
       " 365,\n",
       " 1190,\n",
       " 1219,\n",
       " 932,\n",
       " 1209,\n",
       " 664,\n",
       " 276,\n",
       " 1171,\n",
       " 420,\n",
       " 870,\n",
       " 923,\n",
       " 652,\n",
       " 126,\n",
       " 630,\n",
       " 1134,\n",
       " 232,\n",
       " 326,\n",
       " 594,\n",
       " 154,\n",
       " 845,\n",
       " 607,\n",
       " 649,\n",
       " 383,\n",
       " 1144,\n",
       " 1265,\n",
       " 865,\n",
       " 258,\n",
       " 257,\n",
       " 1235,\n",
       " 397,\n",
       " 568,\n",
       " 944,\n",
       " 395,\n",
       " 158,\n",
       " 587,\n",
       " 869,\n",
       " 597,\n",
       " 141,\n",
       " 645,\n",
       " 1037,\n",
       " 168,\n",
       " 319,\n",
       " 1151,\n",
       " 797,\n",
       " 717,\n",
       " 1189,\n",
       " 819,\n",
       " 285,\n",
       " 1145,\n",
       " 813,\n",
       " 1034,\n",
       " 746,\n",
       " 898,\n",
       " 817,\n",
       " 1276,\n",
       " 1277,\n",
       " 272,\n",
       " 576,\n",
       " 480,\n",
       " 511,\n",
       " 912,\n",
       " 848,\n",
       " 781,\n",
       " 373,\n",
       " 1278,\n",
       " 212,\n",
       " 712,\n",
       " 551,\n",
       " 403,\n",
       " 853,\n",
       " 167,\n",
       " 623,\n",
       " 820,\n",
       " 629,\n",
       " 941,\n",
       " 600,\n",
       " 801,\n",
       " 150,\n",
       " 1162,\n",
       " 707,\n",
       " 1204,\n",
       " 250,\n",
       " 1178,\n",
       " 1210,\n",
       " 1004,\n",
       " 908,\n",
       " 532,\n",
       " 309,\n",
       " 435,\n",
       " 442,\n",
       " 710,\n",
       " 321,\n",
       " 585,\n",
       " 122,\n",
       " 376,\n",
       " 353,\n",
       " 933,\n",
       " 1156,\n",
       " 679,\n",
       " 633,\n",
       " 1221,\n",
       " 130,\n",
       " 1184,\n",
       " 846,\n",
       " 152,\n",
       " 542,\n",
       " 235,\n",
       " 806,\n",
       " 1138,\n",
       " 838,\n",
       " 215,\n",
       " 1000,\n",
       " 897,\n",
       " 727,\n",
       " 182,\n",
       " 764,\n",
       " 763,\n",
       " 734,\n",
       " 313,\n",
       " 750,\n",
       " 852,\n",
       " 290,\n",
       " 308,\n",
       " 400,\n",
       " 741,\n",
       " 202,\n",
       " 193,\n",
       " 280,\n",
       " 184,\n",
       " 926,\n",
       " 1274,\n",
       " 1001,\n",
       " 1199,\n",
       " 667,\n",
       " 816,\n",
       " 835,\n",
       " 1010,\n",
       " 825,\n",
       " 799,\n",
       " 589,\n",
       " 928,\n",
       " 743,\n",
       " 729,\n",
       " 889,\n",
       " 268,\n",
       " 648,\n",
       " 204,\n",
       " 368,\n",
       " 338,\n",
       " 1180,\n",
       " ...]"
      ]
     },
     "execution_count": 172,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sorted(list(range(10000)), key=lambda i: distance(X[random_choose].toarray()[0], \n",
    "                                      X[i].toarray()[0]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "18446744073709551616"
      ]
     },
     "execution_count": 173,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "2**64"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0b10011'"
      ]
     },
     "execution_count": 174,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bin(19)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0b110001'"
      ]
     },
     "execution_count": 175,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bin(49)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0b100110'"
      ]
     },
     "execution_count": 176,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bin(38)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0b100000'"
      ]
     },
     "execution_count": 177,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "bin(49 & 38)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Build Search Engine "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "```\n",
    "Input: Words\n",
    "Output: Documents\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def naive_search(keywords):\n",
    "    news_ids = [i for i, n in enumerate(news_content) if all(w in n for w in keywords)]\n",
    "    # O(D * w) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%timeit\n",
    "naive_search('美军 司令 航母'.split())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "len(news_content)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Input word -> the documents which contain this word"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "transposed_x = X.transpose().toarray()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "word_2_id = vectorized.vocabulary_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "word_2_id['今天']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "id_2_word = {d: w for w, d in word_2_id.items()}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "id_2_word[6195]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "set(np.where(transposed_x[6195])[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "'美军'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "word_2_id['美军']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "word_2_id['司令']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "usa_force = set(np.where(transposed_x[7922])[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "commander = set(np.where(transposed_x[2769])[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "usa_force & commander"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from functools import reduce"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d1, d2, d3 = {1, 2, 3}, {4, 5, 6, 3, 2}, {1, 3, 4}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from operator import and_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "reduce(and_, [d1, d2, d3])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def search_engine(query):\n",
    "    \"\"\"\n",
    "    @query is the searched words, splited by space\n",
    "    @return is the related documents which ranked by tfidf similarity\n",
    "    \"\"\"\n",
    "    words = query.split()\n",
    "    \n",
    "    query_vec = vectorized.transform([' '.join(words)]).toarray()[0]\n",
    "\n",
    "    candidates_ids = [word_2_id[w] for w in words]\n",
    "    \n",
    "    documents_ids = [\n",
    "         set(np.where(transposed_x[_id])[0]) for _id in candidates_ids\n",
    "    ]\n",
    "    \n",
    "    merged_documents = reduce(and_, documents_ids)\n",
    "    # we could know the documents which contain these words\n",
    "    sorted_docuemtns_id = sorted(merged_documents, key=lambda i: distance(query_vec, X[i].toarray()))\n",
    "\n",
    "    return sorted_docuemtns_id"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.where(vectorized.transform(['美联储 加息 次数']).toarray()[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "text = \"\"\"新华社洛杉矶４月８日电（记者黄恒）美国第三舰队８日发布声明说，该舰队下属的“卡尔·文森”航母战斗群当天离开新加坡，改变原定驶往澳大利亚的任务计划，转而北上，前往西太平洋朝鲜半岛附近水域展开行动。\\n　　该舰队网站主页发布的消息说，美军太平洋司令部司令哈里·哈里斯指示“卡尔·文森”航母战斗群向北航行。这一战斗群包括“卡尔·文森”号航空母舰、海军第二航空队、两艘“阿利·伯克”级导弹驱逐舰和一艘“泰孔德罗加”级导弹巡洋舰。\\n　　“卡尔·文森”号航母的母港位于美国加利福尼亚州的圣迭戈，今年１月初前往西太平洋地区执行任务，并参与了日本及韩国的军事演习。\\n　　美国有线电视新闻网援引美国军方官员的话说，“‘卡尔·文森’号此次行动是为了对近期朝鲜的挑衅行为作出回应”。（完）\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "print(text)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import re"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "text = \"\"\"美国有线电视新闻网援引美国军方官员的话说\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pat = r'(新闻|官员)'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "re.compile(pat).sub(repl=\"**\\g<1>**\", string=text)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_query_pat(query):\n",
    "    return re.compile('({})'.format('|'.join(query.split())))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "get_query_pat('美军 司令 航母')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def highlight_keywords(pat, document):\n",
    "    return pat.sub(repl=\"**\\g<1>**\", string=document) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "highlight_keywords(get_query_pat('美军 司令 航母'), content['content'][22987])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from IPython.display import display, Markdown"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def search_engine_with_pretty_print(query):\n",
    "    candidates_ids = search_engine(query)\n",
    "    for i, _id in enumerate(candidates_ids):\n",
    "        title = '## Search Result {}'.format(i)\n",
    "        c = content['content'][_id]\n",
    "        c = highlight_keywords(get_query_pat(query), c)    \n",
    "        \n",
    "        display(Markdown(title + '\\n' + c))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "search_engine_with_pretty_print('春节 假期')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "search_engine()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#%%timeit\n",
    "search_engine('美联储 加息 次数')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "content['content'][2189]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Why ?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "```\n",
    "preprocessing\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## PageRank"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import networkx as nx"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import random"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from string import ascii_uppercase"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "ascii_uppercase"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def genearte_random_website():\n",
    "    return ''.join([random.choice(ascii_uppercase) for _ in range(random.randint(3, 5))]) + '.'  + random.choice(['com', 'cn', 'net'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "genearte_random_website()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "websites = [genearte_random_website() for _ in range(25)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "websites"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "random.sample(websites, 10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "website_connection = {\n",
    "    websites[0]: random.sample(websites, 10),\n",
    "    websites[1]: random.sample(websites, 5),\n",
    "    websites[3]: random.sample(websites, 7),\n",
    "    websites[4]: random.sample(websites, 2),\n",
    "    websites[5]: random.sample(websites, 1),\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "website_network = nx.graph.Graph(website_connection)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.figure(3,figsize=(12,12))\n",
    "nx.draw_networkx(website_network, font_size=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "sorted(nx.pagerank(website_network).items(),key=lambda x: x[1], reverse=True)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:cs231n]",
   "language": "python",
   "name": "conda-env-cs231n-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}