I want to write a basic recommender system in Objective-C, and I'm looking for a basic algorithm to work. Unfortunately, the finished systems are outside the table, since none of them look like for Objective-C.
I will have a database of elements, each with tags (I think films with tags such as horror, action, etc.). Each element would have ~ 5 or so of these tags. When a user first uses the application, their profile will be loaded based on their input into a series of questions, linking some tags to their profile.
As the user continues to use the system and evaluate various elements (based on hate / love / love), I would like to adjust the weighting of recommended tags based on this feedback. I would also like to take a few more properties of my ratings, as their profile grows, for example, the 80s, if it concerns films. Or maybe a director who adheres to the theme of the film.
I prefer to avoid normal (or at least popular) recommendation systems where he looks for similar users to create recommendations. This will have a large item database and minimal users to get started.
Can anyone recommend a good starting point for such an algorithm, I would not want to reinvent the wheel, and there are many?
source
share