Personal recommendation service of Mail.ru
Pulse provides feed based on the user’s interest. Pulse uses 1500 sources base and a machine learning system to select materials for every user. All articles and news are verified in editorial office before going to recommendations section.
Multiple factors contribute to recommendation algorithm, for example, the user’s behavior in social networks and on various websites, popularity of the source, clicks history and many others. The service is constantly learning, so the more user reads, the more precise the recommendations are. Pulse can be used by other media and periodicals, there is a special form for such occasions in the Pulse website.
Other developer projects: