In this project, we have created a recommendation system for restaurants using collaborative filtering (CF). The Yelp Dataset has been used for this project. The general structure of a recommendation system is that there are users and there are items. Users express explicit or implicit preferences toward certain items.
The data: https://www.yelp.com/dataset
Please read about the dataset here: https://www.yelp.com/dataset/documentation/main
Outline of this Project
We wanted to work on these two subtopics for the project:
• Base Model: The base model is to answer the question for the user - "give me more restaurants like this one" Create a database of item-item similarities. Build a CF-based Recommender system using Stochastic Gradient Descent. (70% of the project score)
• Our way to improve the project - Further data analysis on the dataset is done to see if we can make an improvement to the base model created in part 1. This can be adding new features to the dataset to recommend the specific restaurant to users or classifying between two overlapping categories like nightlife and bars or restaurants and fast food or checking how well a new restaurant will do in a certain neighbourhood (30% of the project score)