How does Wikipedia’s “Explore” personalized content discovery feature work

wikipedia

The new Wikipedia app for iOS has an "Explore" feature "designed to make it easier to find random new things to read." The official blog post cites Explore as including the following:

  1. Featured article of the day
  2. Top read articles
  3. Picture of the day
  4. Random articles
  5. Nearby articles
  6. Recommended articles based on what you’ve read

1, 2, and 3 are pretty straightforward.
For #4, over what set is it random? (All articles? Non-stubs? Etc.)
For #5, how does the algorithm determine "nearby?"
For #6, how does it decide what to recommend based on what one has read?

Best Answer

The functionality provided by the Wikipedia App is based on whatever functionality is available on the main Wikipedia site itself.

  • For #4, the random articles functionality is based on Special:Random on the actual site in which an articles in the main namespace is chosen at random. There is some information about this functionality on Wikipedia. However, there is a preference for more complete articles based on what I can infer from their code.
  • For #5, the nearby articles are based on the coordinates of your present location given by your phone and the app will attempt to search for articles with coordinates that are near you. The coordinates of each article is stored in Wikidata, so it would be an easy task for the app to do.
  • For #6, Wikipedia uses an algorithm based on the links and categories of the articles that you have read before to generate a list of articles that may interest you.
Related Topic