HealthyUI Chrome Extension

Chrome Extension for the HealthyUI app with React and Webpack

The purpose of the Bursting the Bubble Chrome extension is to allow users to ”burst the bubble”. That is, we want to allow video viewers to have easier access to information that might challenge their views on politics. However, we do not stop there, and our goal is also to allow these users to easily access information to prove or disprove information found in videos they are consuming. We add another step in our approach by adding a carbon emission tracker to this extension. While users might not necessarily click on the articles we display to them, the extension will display their carbon footprint due to YouTube video consumption, hopefully reducing online data consumption. In fact, it has been shown that people react positively to information bringing awareness to a problem (Wang et al., 2024). Lastly, the end goal of this project is to collect data from users (with their permission) using our extension, which will allow us to conduct the following research regarding filter bubbles. That is, this project was started from the research on the topic of Bursting the Bubble by McGill undergraduates (Wang et al., 2024), as there was a need to collect data regarding filter bubbles and to provide the ever-growing McGillian community a tool to fight against the YouTube recommendation algorithm when necessary. The end goal of this project will be to conduct another research using the data provided from this extension to determine whether students are open to different perspectives. That is, while we cannot enforce the usage of our tool, it will be interesting to study the correlation between watching political videos or important news videos and opening our application to get a new perspective or verify the information given in those videos. The extension works as follows: 1. An article query is initiated by opening the extension on a YouTube video page. 2. A GET request is sent to the backend program, with the video ID as a parameter. 3. Then, our backend program (BP) uses the video ID from the GET request to query the YouTube transcript. It does so by providing the video ID to YouTube’s API. 4. Then, the transcript is processed by assigning weights to each word, which in turn computes the most relevant words for said transcript. 5. This list of words is then used to retrieve articles from the Google Fact-Check API. It does so by providing said list of words in its query, as well as using our personal API key provided by Google. 6. Finally, these articles are sent back to our BP, and then forwarded to the extension itself to be displayed. We note that the anti-siloing part of the extension works the same with some small modifications. That is, instead of fetching articles from the Google Fact-Check API from step 5, we first convert any common 2 word to its antonym, and then use these new queries to fetch articles in the NewsApi database.