What it’s really like to do a hackathon
If you search for “hacker” on Google, you will most likely come across images of people with hoodies, sitting in dark corners working on something that is probably illegal. After participating in a 24-hour hackathon (hacking + marathon), with the aim of solving a difficult data problem, I can assure you that this doesn’t come close to what I experienced.
I am a PhD student at the Vrije Universiteit Amsterdam. My research focuses mainly on news diversity: finding out what kind of news people look at online and how the diversity of news affects them. In my work I often work with digital methods and write programs to analyze data, so I probably always wanted to become a hacker, secretly.
And then this secret dream came true, at least sort of. A few colleagues and I saw an ad for a “media hack day” organized by the Dutch public broadcaster (NPO). The topic was very close to the work of our research group: Designing a recommender system for the online platform NPO start that encurages people to consume more diverse content. A recommender system is an algorithm that selects content for you based on specific rules. We are surrounded by these systems every day: “Top picks for you” on Netflix or “Made for you” on Spotify are examples of such systems. However, most of these algorithms are designed to keep you engaged with the website, no matter what. And if what keeps you engaged with the website is more of the same content (yet another romantic comedy…), then that is what will be recommended to you. Instead of feeding people the same type of recommendations, NPO gave us the challenge to find a way to get people to broaden their view, to get them out of their usual patterns and to consume more diverse content.
As is typical for a hackathon, you sign up as a team (4-5 people) and you get a specific problem to solve with a dataset that comes with it. The challenge: coming up with the best solution in 24 hours (meaning: no sleep!). In this case five teams signed up to compete against each other. Our team (“The Geeky Griffins”) consisted of a wild mix of interdisciplinary scientists from both VU and UvA, ranging from Philosopher to AI specialist. We met via video conferences and were supported by experts from the field. After a lot of discussions, conceptualizing, writing code, and designing interfaces, we gave a 10-minute pitch to a jury of experts and stakeholders at the end of our 24-hour hackathon.
Our idea on how to broaden people’s news horizons involved a “match me up” feature that works a bit like Tinder, but for media consumption. It was great to see what each group came up with in such a short amount of time and to see that we, as academics, can contribute to solving “real-world” problems that affect a lot of people. Hopefully the ideas that arose during this hackathon will be helpful in bringing people to consume content beyond their horizon. It was a cool team experience – and did I mention that we won?
[Big thanks to the whole “Geeky Griffins” team: Nicolas Mattis, Sanne Vrijenhoek, Marijn Sax and Myrthe Reuver]