This live twitter sentiment analysis marks a tweet as positive or negative by running several machine learning classifiers together.

All the machine learning classifiers cast their votes on a particular tweet and then the tweet's confidence level is marked.

• Install Natural Language Processing Toolkit (nltk)

import nltk
nltk.download()

This will download all of the corpora.

• Install tweepy using pip

pip install tweepy

• Generate Consumer key, Consumer Secret, Access token and Access secret by registering an application on Twiiter Apps. Copy and paste these required keys in twitter_sentiment_analysis.py file.

• Make a new directory pickled_algos in the root directory.

• Run module_for_sentiment_analysis.py file to train machine learning models on movie review texts.

• Run twitter_sentiment_analysis.py to stream live tweets on console and mark them as positive or negative.

• Simultaneously run graphing_live_tweets.py to generate live graph for the marked categories.

twitter_sentiment_analysis/
    README.md
    graphing_live_tweets.py
    module_for_sentiment_analysis.py
    sentiment_mod.py
    twitter_sentiment_analysis.py
    .gitignore
    short_reviews/
        negative.txt
        positive.txt

If there is any issue in the source code, send me pull request and contribute to this project.