Restaurant Recommender System Using User-Based Collaborative Filtering Approach: A Case Study at Bandung Raya Region
Alif Azhar Fakhri, Z. K. A. Baizal, Erwin Budi Setiawan · Journal of Physics Conference Series · 2019
Culinary becomes one of the needs of today's society. The large number of restaurant choices and also lack of information about the restaurant become an obstacle to people's needs in choosing a restaurant. In this paper, we build a recommender system that can recommend the restaurant in Bandung area. However, today, users want to get a restaurant with a good reputation and fit their tastes, so that restaurant ratings from other users are required in the restaurant recommendation process. We implement a user-based collaborative filtering method for recommend a restaurant personally, based on ratings given by other users. We also implement two similarities, i.e., user rating similarity and user attribute similarity to find the proximity between users. We use Mean Absolute Error (MAE) to evaluate accuration of rating prediction. The best MAE result of each performance is 1.492 for calculation without user attributes and 2.166 for calculation with user attributes.