This repository contains all the data and code necessary to reproduce Study 1 and Study 3 in the paper.
The abstracts for the papers are cscw50_papers-abstracts.csv
The "gold-standard" known analogy pairs are enumerated in gold_analogy_pairs.txt and explained in gold_analogy_pairs_with-explanations.csv.
For convenience for reviewers with Unix-like systems (e.g., Linux, OS X), the full experiment can be run from start (assuming annotations are already obtained) to finish (comparing precision of matches) by running run-pipeline.sh.
Note: the code in this repository is in Python 2
Python module dependencies (note: most of these can be obtained by simply downloading the Anaconda distribution)
- Pandas
- NLTK
- Numpy
- Scipy
You'll also need gensim-formatted base vectors (will make more sense after inspecting get-vectors.py), placed in the models/ subdirectory. The vectors used for Study 1 and 3 can be downloaded here: https://www.dropbox.com/s/7bqk9i2fd56xchk/glove.papers.600d.cscw%2Bchi.2010-2017.txt?dl=0.