For our Data Science Discovery Project with Activism Always, we focused on creating a platform based off of Twitter data. We were tasked with improving and building upon the code left by previous interns, which provided a method of building a tree data structure, where the nodes are based on popular hashtags(ex: #BLM), the children are hashtags that are related, and the edge weights represent how closely related the parents and children hashtags are. Using this, we further optimized the algorithm, improving its runtime and connecting it to our new DynamoDB database. Additionally, we took in data from Twitter’s API to perform sentiment analysis. Finally, we used Twitter data to analyze the data of individual profiles and to output relevant statistics for each profile.

Project Report

Term
Spring 2022
Topic
Social Sciences
Technical Area(s)
Machine Learning (ML)