Twitter Posts Political Ideology Classification (Student Presentation, Group 15)

This is a STAT 451 class project presentation by Han Cao, Siyi He, Qiwen Zeng This presentation is shared with the students' permission. Abstract: Some reports[5] have been indicating that social media has a strong impact on political ideologies of the society. Emerging conspiracy theories suggest that other countries have been using social media platforms to manipulate presidential election of the United States. However, in order to massively monitor such behavior online involving technologies that enable computer to understand the political ideologies behind the texts. We proposed using machine learning algorithms to solve such a problem. We experimented Support Vector Machine, XGBoost, and Multinomial Naive Bayes on a data set collected from twitter. We eventually achieved an accuracy of 76%.