Multi-Criteria Ranking: Next Generation of Multi-Criteria Recommendation Framework
Yong Zheng, David Wang
Abstract
Recommender systems have been served to assist decision making by recommending a list of items to the end users. Multi-criteria recommender system (MCRS) is a type of recommender systems which enhance recommendation performance by taking user preferences on multiple criteria. Traditional algorithms for MCRS usually predict user ratings on these criteria, and finally estimate the overall rating by different aggregation functions. In this paper, we propose a new multi-criteria recommendation framework in which we can take advantage of Pareto ranking based estimated preferences on multiple criteria, and infer a ranking score for top-<i>N</i> recommendations. The proposed framework is general enough and all existing algorithms in MCRS can be reused to be integrated with our framework. We demonstrate the effectiveness of the proposed framework by evaluating top-<i>N</i> recommendations over four real-world data sets.