The goal of the project is to predict possibly interesting subreddits to a user based on their comment history. The hypothesis of the recommended model is, given an ordered sequence of user subreddit interactions, patterns will emerge that favour the discovery of particular new subreddits given that historical user interaction sequence.
Implemented three models :
- Simple Popularity Based Recommender System.
- Item-Item Based Collaborative Filtering Recommender System.
- User Based Collaborative Filtering Recommender System