What's trending on Wikipedia?

by on under Research
1 minute read

There are as many opinions as there are people, they say. After a recent study we have done in LTS2, I can paraphrase it. There are as many opinions as there are languages on Wikipedia. In our case though it was more about people’s interests rather than opinions.

Trends distribution across languages

In a recent study, we found that interests of Wikipedia readers largely depend on the language in which they read Wikipedia. We compared English, French, and Russian language editions and found that certain topics are more popular in one language than in the others. For instance, topic Sports dominates among users reading Wikipedia in English, Francophone readers prefer Wikipedia articles about Movies, and Russian-speaking audience is obsessed with Wikipedia articles related to Science. Also, we found that topics related to pop-culture (e.g. release of a movie by Marvell) and global dramatic events (September 11) are popular across all languages that we have studied.

Comparison across languages

So, what’s trending on Wikipedia? It depends. Specifically, it depends on the language you are interested in. If you want to learn more, watch our presentation at Wikipedia Workshop 2020 and read a review of our study by Isaac Johnson in Wikipedia Research Newsletter. If you are interested in the technical details, our method, and the dataset, take a look at the paper.

I worked on this project with Joëlle Hanna, Nicolas Aspert, Benjamin Ricaud, and Pierre Vandergheynst. You can find this and many other Wikipedia-related projects and resources by our lab on Wikipedia Insights project website, for instance, an anomaly detection algorithm and a graph-based dataset with pre-processing tools that we used to detect trending topics on Wikipeida.

If you are interested in the graph visualization tools we used in this project, take a look at the following tutorials:

Stay tuned for more!

Collective behavior, Data analysis, Graph, Time-series, Dynamic Networks, Machine Learning, Research, Network analysis, Wikipedia
comments powered by Disqus