What we can Learn about Russian IRA Trolls during 2016 U.S. Election from Twitter (DSI-SRP)
This DSI-SRP fellowship funded Catherine (Nayeon) Kim to work in the laboratory of Professor Jennifer Larson in the Department of Political Science during the summer of 2019. Catherine graduated in 2020 with majors in Mathematics and English Creative Writing.
The project funded by this fellowship aimed to study the impact of online propaganda on public opinions. After reading research literature and journalistic investigation pieces, Kim decided to perform analysis on the two datasets released in 2018 by Twitter regarding those tweets from “identified” Russian trolls who attempted to manipulate the public with fake news and strong expression of emotions during the 2016 election. The datasets contained all tweets created by 3608 Russian troll accounts.
This project was Kim’s first experience with a large dataset. Thanks to the data visualization and analysis workshops offered by DSI, she was able to better utilize Python data libraries to extract information from the datasets. She studied how those accounts interacted with one another (e.g. retweets, replies), the languages in which the accounts tweets, the attention those accounts’ activities drew from non-IRA-identified accounts. Additionally, she analyzed how those accounts participated in conversations regarding 2016 presidential candidates. Kim was surprised to find that they didn’t always say great things about Trump, as many of us might’ve expected. She also found how the datasets provided by Twitter didn’t always contain accurate information. The language identifications were wrong for many non-English tweets.
There is much more information to explore, and can be extracted from the datasets. Kim stated that this project showed her the power of data science and how programming language and libraries can be utilized to quickly extract multi-dimensional information from datasets. Her experience with analyzing the dataset also showed her that producing good data analytics results require skills, patience, and diligence. Kim says she learned that, “we can’t just use “data science” to magically turn datasets into conclusions; solid data science skills and careful analysis are imperative for identifying helpful, reliable information.” The Data Science Institute (DSI) and her mentor, Dr. Larson, supported Kim through their great mentorship, helping her to make progress on her the project. This summer research experience exposed Catherine Kim to the power and potential of data science and strengthened her interest in computational research.
In addition to receiving support through a DSI-SRP fellowship, this project was supported and facilitated by the DSI Data Science Team through their regular summer workshops and demo sessions.