Problem:
Analyze tweets from Republican party candidates during the 2016 presidential primaries to see what kinds of tweets generated the most response from the public.
Dataset:
We will mine tweets by using the Python library Tweepy to connect to the Twitter API. The data that is returned from the Twitter API is in JSON format. The data for each tweet contains the text content of the tweet as well as metadata, including the number of retweets and favorites.
Proposed Solution and Real world Application :
Our proposed solution is to obtain both a sentiment analysis score and list of “trigger” words for each tweet. “Trigger” words refer to monograms or bigrams that influence the sentiment of tweets. Each tweet will be given a score based on the percentage of followers that commented/retweeted/favorited that tweet. We aim to visualize the relationship between the content of the tweet and its public reception. This solution can be used to understand the type of rhetoric that is successful in garnering a large public response, whether it be positive or negative. In addition to politicians, the framework can be applied to other public figures such as business leaders and athletes, in order to determine the effectiveness of their tweets.
About this project:
This is our group project of ECE 143 Programming for Data Analysis. Our group member including:
- Jared Johann Leitner
- Haonan Song
- Yuchen Tang
- Louise Xu
Thank all members above for our contribution for this project, well done!
Our GitHub repository url is below:
Yuchen Tang 