We've worked on social media projects before, trying to understand the connection between social media and virality. This time, we wanted to see how those themes applied to one of society's most pressing issues today.
Furthermore, Max has done machine learning research on the effects of online moral outrage and hate speech, so this project was an interesting continuation of that work in a new field.
What it does
Our project tells the narrative of social media and the interplay between hate and counterhate using interactive data visualizations.
How we built it
We built the webpage using Python and ReactJS.
Python was used to clean and analyze the twitter dataset provided by Georgia Tech, ultimately generating JSON files containing time series and geolocation data related to hate and counter hate tweets. We also used Python’s Selenium library to scrape GoFundMe for data on its official #StopAsianHate funds. Once again, a JSON file was generated with this data.
After the data analysis segment of this project, we used ReactJS to develop a website to house our interactive data visualizations and other important information regarding the subject of asian hate. Specifically, the data visualizations were created using the nivo and/or react-usa-map libraries, as well as Adobe Illustrator.
Challenges we ran into
Besides the time series, geolocation, and GoFundMe data, we were also originally interested in looking at the spread of hate and counter hate tweets within the context of a social network. The Georgia Tech Research group provided a huge dataset with 489K nodes and over 700 million edges. Unfortunately, this dataset was simply too sparse and big for us to process and analyze. Furthermore, we wanted to look at the historical performance of various hash tags on Twitter, but unfortunately, could not get access to Twitter’s API.
Accomplishments that we're proud of
Doing research and telling an honest and interesting narrative about a topic as complex and important as asian hate. Also, we were able to accomplish a surprising amount of work in such a short time period, and we have even more time after this Vizathon to continue polishing our project.
What we learned
We learned how to conduct even more powerful web scrapes with Selenium instead of Beautiful Soup. In hopes of educating others about this important topic, we ourselves learned a lot about the role of social media in the spread of anti-Asian hate.
What's next for Social Media: Agent of Transmission
We will be getting our own custom domain to publish the site on in order to push out this info to more people. We also plan on tackling the social network and hashtag problems after this Vizathon considering that we will have more time to think of better solutions.