/Rating-Prediction-using-RNNs

Final course project for CSE 258 (Fall 2022; Prof. Julian McAuley; UC San Diego)

Primary LanguageJupyter Notebook

Rating Prediction using RNNs

CSE-258-Project

This repository contains our final course project for CSE258: Web Mining and Recommender Systems (Fall 2022; Prof. Julian McAuley; UC San Diego).

The problem of predicting the rating a user will assign to an item has become an interesting problem in recommendation systems that use these predictions to recommend new products to users. This problem is called Rating Prediction, and in this project, we build various models which use different features, for e.g. user-item interactions and reviews, to predict the rating a user will give to an item. We explore three major classes of methods:

  1. Content-based Filtering Methods
  2. Collaborative Filtering Methods (Latent Factor Models)
  3. Text-based Methods (Bag-of-Words, TFIDF, RNNs)

All the implementation details and results are enclosed in the report.