Twitter Sentiment Analysis and Prediction

Posted on | In

Just few development blog about twitter data analysis, following the tutorials on Kaggle. Complete code and downloadable files at Github repo

Using Jupyter Notebook

Tweets.csv (basically tweets about U.S airlines) also can be found at repo.

Twitter sentiment

  1. read tweets

  1. calculate frequency of polarity

Twitter prediction

  1. import basic library

  2. read data from tweets csv file

  3. frequency of columns

  4. find general moods for airlines

  5. compare moods of different airlines

  1. figure out reasons for negative tweets

  2. compare the negative reasons proportion of different airlines

  3. wordcloud for negative reasons

  4. accuracy of different classifiers performance