A configuration-based recommender system for supporting e-commerce decisions

详细信息    查看全文

• 作者：Michael Scholz ; ^a ; ^{michael.scholz@uni-passau.de} ; Verena Dorner^b ; ^{verena.dorner@kit.edu} ; Guido Schryen^c ; ^{guido.schryen@wiwi.uni-regensburg.de} ; Alexander Benlian^d ; ^{benlian@ise.tu-darmstadt.de}
• 关键词：E-commerce ; Recommender system ; Attribute weights ; Configuration system ; Decision support
• 刊名：European Journal of Operational Research
• 出版年：2017
• 出版时间：16 May 2017
• 年：2017
• 卷：259
• 期：1
• 页码：205-215
• 全文大小：861 K
• 卷排序：259

文摘

Attribute weighting methods used in recommenders do not fit to shopping tasks. We present a new attribute weighting method that naturally fits to shopping tasks. The novel method is based on a product configuration system. An experiment shows that our method is suitable to be used in shopping tasks. We found that our method and SWING produce rather similar attribute weight estimates.