mRHR: A Modified Reciprocal Hit Rank Metric for Ranking Evaluation of Multiple Preferences in Top-N Recommender Systems


Peker S., KOÇYİĞİT A.

17th International Conference on Artificial Intelligence - Methodology, Systems and Applications (AIMSA), Varna, Bulgaria, 7 - 10 September 2016, vol.9883, pp.320-329 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Volume: 9883
  • Doi Number: 10.1007/978-3-319-44748-3_31
  • City: Varna
  • Country: Bulgaria
  • Page Numbers: pp.320-329

Abstract

Average reciprocal hit rank (ARHR) is a commonly used metric for ranking evaluation of top-n recommender systems. However, it suffers from an important shortcoming that it cannot be applied when the user has multiple preferences at a time. In order to overcome this problem, a modified version of ARHR metric is introduced and applied to grocery shopping domain by conducting a series of experiments on real-life data. The results show that the proposed measure is feasible for ranking evaluation of Top-N recommender systems in the cases where the users have multiple preferences at a time or a specific time interval.