基于多样性选择因子的多目标混合推荐研究  

Research on Multi-objective Hybrid Recommendation Based on Diversity Selection Factor

作　　者：李冰[1] 王虎[2] 王锐[3] LI Bing;WANG Hu;WANG Rui

机构地区：[1]江西师范大学软件学院,江西南昌330022 [2]武汉理工大学管理学院,湖北武汉430070 [3]江西理工大学经济管理学院,江西赣州341000

出　　处：《武汉理工大学学报（信息与管理工程版）》2019年第3期299-304,共6页Journal of Wuhan University of Technology：Information & Management Engineering

基　　金：国家自然科学基金项目(71661013);江西省教育厅科学技术研究基金项目(150320);江西师范大学青年成长基金项目(12017193);江西师范大学博士启动基金项目(12016422)

摘　　要：目前的服务推荐领域主要偏向于研究提升推荐的精准度,缺少对多样化推荐方面的研究。针对这一现状,提出一种考虑多样化选择因子的混合推荐策略。首先,提出客户对服务项目的多样性选择因子,具体包括二次选择因子、个性选择因子和热门选择因子。其次,基于3个多样性选择因子,提出一种多目标的混合推荐模型,在实施服务推荐之前先挖掘客户对不同多样化因子的偏好程度,运用马尔科夫链计算不同因子的权重,基于混合算法得出能同时满足客户所有多样性选择因子的服务项目。最后,针对冷启动问题,采用多智能体模拟的方法确定了冷启动阶段3个多样性选择因子的搭配比例。所提出的推荐策略弥补了已有的服务推荐理论的不足,对未来的智能化、主动化电子商务服务平台的设计提供了借鉴。The current service recommendation field mainly prefers to study the accuracy of the promotion recommendation and lacks research on the diversity recommendation. In view of this situation, this paper proposes a hybrid recommendation strategy which takes into account the diversity of selection factors. Firstly, the diversity selection factors of service items are proposed, including secondary selection factors, personality selection factors and popular selection factors. Then, based on three diversity selection factors, a multi-objective hybrid recommendation model is proposed. Before implementing service recommendation, the customer′s preference for different diversity factors is first discovered, and the weight of different factors is calculated using the Markov chain. Based on the hybrid algorithm, the service items can satisfy all the diversity selection factors of customers at the same time. Finally, the matching ratio of three diversity selection factors in cold start-up stage is determined by using multiagent simulation. The recommendation strategy put forward in this paper makes up for the shortcomings of the existing service recommendation theory and provides reference for the future intelligent and active e-commerce service platform design.

关 键 词：服务推荐 多样性因子 混合推荐 推荐模型 个性推荐 

分 类 号：F272[经济管理—企业管理] TP301[经济管理—国民经济]