ProbQA interactive recommendation system for games

ProbQA interactive recommendation engine helps to find the next game to play if you don't know the keywords. It's a rare occasion when users know in advance the keywords to search for to find the next game to play. To know the keywords, users have to talk about games or read the gaming news. Though this is the case for some users, the others have only a rough idea on what they would like to play next. For such cases a recommendation engine can be helpful. On one hand, recommendation engine could recommend similar games to the ones the user has already played. But for some users it may be boring to play clones of a game. They rather like games based on features present in them. Or even want to try something different. So how to solve such a problem? One way is to interview the user: to ask him/her questions about the next game to play. This also solves another problem: users may like games which are not usually rated as "top 10 games" in the articles. Though this applies to minority of users, because the majority of users like the mainstream top games, that's why those games are that popular. Both problems are solved by a recommendation engine that lists top games in each step according to their probabilities given the answers to the questions in the previous steps. ProbQA is one such engine. It collects statistics about preferences of the users given each quiz. So it constantly learns. This kind of Artificial Intelligence can be applied to many other targets than games, e.g. products and services, films and music, etc. We decided to start with games here: http://probqa.com/