A Product Recommendation Algorithm Based on Knapsack Optimization

摘要：

Personalized Recommender System becomes an important research field in Electronic Commerce, and the main goal of current recommendation models is provide Best-Service to users. But, from enterprises viewpoint, the Max-Earning strategy is necessary to improve the benefit of enterprise. To solving this problem, knapsack model is applied to describe the commonly used Top-N recommend mechanism firstly. Then, the enterprises earnings are described as a constraint in knapsack model, a product recommended algorithm is proposed at the basis of optimization of knapsack problem. Experimental results show the proposed algorithm has similar performance with CF model when earning requirement and amount of recommended products is lower. So, both users value and enterprises value are improved through the proposed algorithm.

关键词： Recommendation System Collaborative Filtering Knapsack Problem Best-Service strategy Max-Earning strategy

作者: Linqi Gao

作者单位: College of Management, Tianjin Normal University, China

会议类型: 国际会议

会议名称: The Eleventh Wuhan International Conference on E-Business(第十一届武汉电子商务国际会议)

会议地点: 武汉

会议语种:英文

页码: 38-42

在线出版日期: 2012-05-26（万方平台首次上网日期，不代表论文的发表时间）