A new framework for collecting implicit user feedback for movie and video recommender system

Abstract

In today’s digital world due to unlimited content, product and services available online, finding an item that satisfies user requirement and taste by simply web searching is near impossible. Recommender systems are information filtering tool which provides personalized results. Movie and video recommender system is also gaining popularity due to the growth in online streaming video content Web sites and its subscriber. Accuracy and efficiency are two major aspects of a recommendation engine because it is directly related to user experience. To achieve higher accuracy, user feedback is required which can be collected either explicitly or implicitly. Explicit feedback is not always available and not always unbiased, so implicit feedback seems to be a better option for user preference collection. In this paper, a new framework is proposed which collects the implicit user feedback (along with explicit) for a movie and video recommender system. Implicit feedbacks can be converted to explicit feedback using the proposed UARCA which can be used to improve the accuracy of recommendation engine.