Reader comment analysis using the New York Times (c) community API
- Requires python 2.7, numpy, matplotlib, scikit-learn and xmltodict
- Rename config.example to config.py and edit in your API keys
- Load from apiQuery import TimesComments, then TimesComments('20140105',False) to query the NYT API for comments from January 5th, 2014 (or choose another date)
- Load from classifyData import *, then classifyData('20140105') to begin classifying comments and training the learning model
- With learning model trained, from commentAnalysis import CommentAnalysis, then CommentAnalysis('20140105') to find poems and calculate word frequencies
Use apiQuery.py to load new comments from the NYT API
Use classifyData.py to train the learning algorithm
Use commentAnalysis.py to perform analysis on comments
There is a blog post with more details about the ideas behind this project