Playtime - based vs Price-based Rating in Video Games Recommender System

摘要

Recommender System in one of the solutions to filter information presented to user to avoid information overloading. On video games e-commerce platform, recommender system was implemented make time taken to choose games to buy and play relatively shorter. Researches about video recommender system mostly focused on duration of playtime to predict user preferences towards certain games. In this paper we proposed to broaden the parameters by including price-based rating. Video games recommender system is developed, and price-based ratings were feed into the recommender system to see the impact of the rating on the recommender system performance. Experiments to compare recommender system performance using playtime-based rating and price-based rating were executed and it is found that in terms of rating prediction, playtime-based rating shows better performance but in terms of hit rate, price-based is better.