Date of Publication :22nd February 2018

Abstract: In an era of information age, recommender system helps users to make an effective decision. Collaborative filtering is one of the techniques to provide a personalized recommendation to users. Collaborative filtering based recommender technique provides the recommendation by aggregating ratings from similar users to predict ratings for an active user (who wants a recommendation). The similarity has a greater impact because it acts as a criterion to identify a group of similar users whose ratings will be merged to generate a recommendation for the new item for an active user. However, there are a lot of issues in Collaborative filtering for e.g. data sparsity and cold start, which can be removed by incorporating trust information. We propose a methodology to include temporal context information in providing accurate rating prediction along with Trust matrix and also propose a framework to analyze the performance of Trust-based recommender algorithms on Film Trust dataset which includes temporal context information.

Reference :