Recommendation in online shopping malls: Results and experiences

Abstract

Some portal sites have been providing price-comparison services. For the price-comparison services, however, they do not provide personalized recommendations. This paper proposes a strategy of recommendations for the price-comparison services. The strategy uses the click-log data to identify user's preference. We apply the strategy to three recommendation methods. Through the experiments, we have shown the possibility of recommendation for price-comparison services with the real-world data.