presidential bias predictor
Top 5 Course Projects
Project Date: December 2020
Technologies: Python, Unofficial TikTok API, Porter Stemmer Algorithm, Gaussian Classifier, Naive Bayes Algorithm, Google Colab
Role: Data Wrangling, Datasets, API Handle
Team: Elias Arghand, Elias Fang, Norman Ta, Wesley Wang, Vivian Wu
Can we create a classification model to predict the 2020 presidential bias of politically-focused TikTok videos?
Overview
TikTok, a social media platform that recently rose to fame around the world, has set the
stage for users to express their political ideologies in short videos with the recent 2020 Presidential Election. We
decided to gather data leveraging various political hashtags to see if we can predict the presidential bias of a
politically-focused TikTok video.
Through the predictive model, we generated with the data we scraped from the most popular TikTok videos, we were able to accurately predict the bias of any given TikTok video and also discovered that TikTok has a predominantly left demographic. The findings of our project has allowed us to better understand the role of machine learning in predictive analysis and our data can also be used to better understand the biases that exist in social media content.
Through the predictive model, we generated with the data we scraped from the most popular TikTok videos, we were able to accurately predict the bias of any given TikTok video and also discovered that TikTok has a predominantly left demographic. The findings of our project has allowed us to better understand the role of machine learning in predictive analysis and our data can also be used to better understand the biases that exist in social media content.