Group Recommender System Based on Rank Aggregation - An Evolutionary Approach

摘要

Recommender systems (RSs) have emerged as a remarkable tool that very effectively handle information overload problem caused by unprecedented growth of resources available on the www. RSs research has mainly focused on algorithms for recommending items for individual users. However, Group recommender systems (GRSs) provide recommendations to group of persons i.e. they take all individual group members' preferences into account and try to satisfy them optimally. The well known Kemeny optimal aggregation generates an aggregated list that minimizes the average Kendall tau Distance from the input lists; however such aggregation is NP-Hard. In this work, we design and develop a novel approach to GRS based on Kemeny optimal aggregation using genetic algorithm (GA). We have employed edge recombination operator (ERO) and scramble sub-list mutation as genetic sequencing operators. Experimental results clearly demonstrate that proposed GA approach to rank aggregation (RA) based GRS, GA-RA-GRS outperforms the well known GRS techniques.