Abstract

Many recommender systems suffer from the popularity bias problem: popularitems are being recommended frequently while less popular, niche products, arerecommended rarely if not at all. However, those ignored products are exactlythe products that businesses need to find customers for and theirrecommendations would be more beneficial. In this paper, we examine an itemweighting approach to improve long-tail recommendation. Our approach works as asimple yet powerful add-on to existing recommendation algorithms for making atunable trade-off between accuracy and long-tail coverage.

 

Quick Read (beta)