- Note: don’t forget to install dependencies and establish app configs
##pip install list:
- tweepy
- nltk
- pymongo
##Obtaining a corpus of keywords The corpus of keywords used to filter tweets from the Twitter streaming API was obtained by:
- Creating a list of the top 25 most popular American politicians on Twitter (by follower count)
- Obtaining the tweet history of each politician since the beginning of 2015 (1/01/2015)
- Extracting a list of n-grams (unigrams, bigrams, and trigrams) from the body of these tweets
- Filtering out n-grams irrelevant to politics
##File structure (what does what)
####credentials.py Dependency file contains the credentials for your twitter application. You must create and register an application at Twitter (see the developers page) in order to get the 4 keys necessary to operate the tools
####db.py Test script to start connection with local MongoDB.
####nltk_politicians.py A script that will grab scrape politician twitter accounts for the past week and write a list of the tweets in ...
####politician_text.txt Text file containing the politicians tweets that were scraped in nltk_politicians
####politix.json Junk -- streaming Twitter json objects saved in a native JSON file. Useful until mongodb database is set up and universally accessible.
####test.py Script that opens up twitter streaming API and handles each tweet -- writes to politix.json but will be writing to local mongoDB
####tweets_to_CLI.py Copy of test.py but does not write to DB. Instead the script prints the output to command line interface.