Group 9: Political Sentiment Analysis on Twitter

Team Members

• Akash Boghani - aboghani@ucsd.edu
• Khushboo Agrawal - khagrawa@ucsd.edu
• Lina Yi - lkyi@ucsd.edu
• Haoran Wu - haw004@ucsd.edu

Objective

Twitter is the modern battleground of politics. This project aims to perform sentiment analysis on the tweets of popular politicians, and find interesting correlations based on:

• Their tweeting habits
• Their stance on major issues
• Their support base

Requirements

In order to run the code in this repository, you will need to install the following dependencies:

• tweepy: pip install tweepy
• pandas: pip install pandas
• numpy: pip install numpy
• textblob: pip install textblob
• vader: pip install vaderSentiment
• wordcloud: pip install wordcloud
• matplotlib pip install matplotlib
• plotly pip install plotly
• plotly-express pip install plotly-express
• statistics pip install statistics
• sklearn pip install sklearn

File Structure

Presentation

• Twitter Sentiment Analysis.pdf This is the final presentation converted to pdf format for your reference.

Demo

• Code_Demo.ipynb This Jupyter notebook contains the demo. To run, clone the repository. Then, navigate to the repository in Terminal, and type jupyter notebook. Click on Code_demo.ipynb to run the demo.

Data Collection

• collect_data.py This python code file fetches the tweets given the twitter handle of the politician and stores in a csv file.

Data Processing

• process_tweets.py This python code file is used to process the tweets from the csv file by cleaning the data and adding the subjectivity and polarity.

Data Visualization

• pie_scatter.py This python file contains the functions to plot:

  • Scatterplot of the average polarity and subjectivity of the tweets of the politicians.
  • The K-Nearest Neighbors plot for finding out the boundary between the republicans and democrats.
  • Functions to plot the pie charts of percentage polarity for each politicians.

• gen_wordcloud.py This python file contains the functions to generate the word cloud for positive and negative polarity of the tweets of politicians.

• engagement_plot.py This python file contains functions for plotting the engagement vs polarity bubble chart for each politician.

• average_tweets.py This python file contains functions for plotting the average tweets per day of the politicians

• christchurch_shooting.py This python file contains functions for plotting the comparison of the sentiment polarity of tweets with the average tweets during the christchurch shooting.

Logistics and Data

• twitter_credentials.json This python file contains the configuration of the twitter developer app for extracting the tweets.

• all_handles.txt This text file contains all the twitter handles of the politicians for which we have extracted the data and analyzed.

• real_names.txt This text file contains the name of the politician real name and their twitter handle

• data/ folder The folder contains all the csv files of the politicians containing the tweets and other attributes