Bored of COVID-19? We were, too. Because of this, a bunch of DPhil students from our research group met online on the weekend of 24/25 April 2021 for the first ever “HCC Hackathon”. We aimed to create a software artifact together, and have fun whilst doing so. We ended up pursuing four, related projects over this weekend.
YouTube survey: Reviewing, adapting and extending on similar surveys, we created a draft for a survey to gauge how well YouTube’s recommendation algorithm seems to reflect users’ interests and primary (long-term) goals for using YouTube, as well as the steps they take (if any) to improve their recommendations.
YouTube watch history: Our aim with this project was to offer users the ability to more intuitively control their YouTube recommendations so as to align with/reinforce their long-term goals/interests (and not reinforce their short-term desires). This takes the form of an extension that serves as a “privacy mode” toggle, which allows users to stay logged in but decide when they want to see more of the same content they watch and when they don’t.
YouTube watch history analysis: As part of the goal of enabling users with more choices over their YouTube recommendations, we would like to offer the users the ability to analyse and visualise their own watching histories (maybe integrate this functionality into our recommender as well).
Harms data collection tool: Online harms are widespread, so we built a tool that allows users to report harms that they come across to us, via a browser extension.