Every article is identified by a token
(for example 246311
). For every article we have a JSON file that contains
many recommendations for that article, based on 3 different algorithms: logs, content, and myad
- The logs algorithm finds articles visited by users that also visited the current article.
- The content algorithm finds articles that contains similar text to the current article.
- The myad algorithm finds articles that have been bookmarked within the same folder that the current article.