Skip to content

faychu/FDHNE

Repository files navigation

videolist.pickle: 类型set,共74088条,即有视觉的video id
video_with_vis&txt.pickle:类型list,共44144条,即具有视觉和文本特征的video id
relate_pair.pickle:相关关系对,类型set,例如((1,2),...),里面的视频id均有视觉和文本特征,共84214条,正反均有
video_rel.pickle:具有relate关系的视频(且均有视觉和文本特征),共31362条
relate_set_all.pickle:相关关系对,类型set,例如((1,2),...),里面的视频id均有视觉和文本特征,共116960条,正反均有
video_rel_all.pickle:具有relate关系的视频(且均有视觉和文本特征),共43864条
up_pair_all.pickle:类型set,(vid, authorid),针对具有视觉特征的video,共74088条
author_all.pickle:类型set,针对具有视觉特征的video,共34460个author

up_pair.pickle:类型set,(vid, authorid),针对具有视觉及文本特征的video,共44144条
author.pickle:类型set,针对具有视觉及文本特征的video,共20195个author

video2label_vis&txt.pickle:类型dict,{原始vid:原始label}


标号字典:
vid_dict.pickle:类型dict,(原始id:新id(0-44143))
aid_dict.pickle:类型dict,(原始id:新id(44144-64338))
lid_dict.pickle:类型dict,(原始id:新id(64339-64356))


特征:
hlf_vis&txt.pickle:类型dict,针对具有视觉及文本特征的video的hlf特征,key值范围(0-44143)
tfidf.pickle:类型dict,针对针对具有视觉及文本特征的video的tfidf特征,key值范围(0-44143)

Context:
context.pickle:类型dict,{vid:{0:lid,1:[vid],2:aid}}

输入模型中只需要:
hlf_vis&txt.pickle
tfidf.pickle
context.pickle

类别信息:
Counter({
'Entertainment': 7462,
'News&Politics': 4219,
'Comedy': 4122,
'Music': 3988,
'People&Blogs': 3107,
'Howto&Style': 3076,
'Gaming': 2893,
'Sports': 2710,
'Science&Technology': 2449,
'Autos&Vehicles': 2398,
'Pets&Animals': 1918,
'Film&Animation': 1911,
'Education': 1439,
'Nonprofits&Activism': 1183,
'Travel&Events': 1164,
'Shows': 93,
'Trailers': 7,
'Movies': 5})

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages