Recommendation Engine with various functions to get recommendations.
User can use any one of the functions they prefer to get recommendations. I have already done used one technique to get a recommendation for the 56th item in the data set in recommendationengine.py file. You can replace it with your desired functions also to get recommendations.
MovieLens Dataset is used.MovieLens data sets were collected by the GroupLens Research Project at the University of Minnesota.This data set consists of: -> 100,000 ratings (1-5) from 943 users on 1682 movies. -> Each user has rated at least 20 movies. -> Simple demographic info for the users (age, gender, occupation, zip).
MovieLens Dataset is attatched in this repository.
For deepdiving in to various recommendation techniques and concepts, refer to book named Programming Collective Intelligence by Toby Segaran.
Depending on the speed of your computer, you may notice a pause when getting recommendations this way. This is because you’re working with a much larger dataset. The more users you have, the longer user-based recommendations will take.