点击次数：

论文类型：会议论文

发表时间：2015-01-01

收录刊物：CPCI-S

卷号：124

页面范围：1593-1598

关键字：fuzzy theory; e-commerce; personalized recommendation; kansei image

摘要：Personalized recommendation technique is an important technology to solve information overloaded in e-commerce. But recently new commodities are emerging constantly, so it is required to recommend commodities that consumers aren't familiar with but interested in. Consequently our study proposes a method to mine consumer's preference from the respective of consumer psychology. Consumers' affective needs can be described in the form of kansei image preference and kansei image weight. Then recommendation results are produced to meet consumer's quantitative affective needs. Finally by real historical data of 15 consumers online and surveys in the example of garments, validity of the method is verified.