| Summary: | 碩士 === 中國文化大學 === 資訊管理研究所 === 95 === The rapid development of internet has changed the patterns of our life. Something that only can be accomplished in the real world in the past can be realized via the internet nowadays, one of them is consumer’s behavior. Users of the internet can acquire a huge amount of information easily, but not all of that is what they really need. It not only reduces the benefit of the companies, carries out the marketing strategy inefficiently, loses consumers’ loyalty but wastes the cost as a result of excessively providing useless information. On the part of consumers, filtering abundant information wastes time and decreases the convenience of internet shopping. Avoiding situations like that, companies provide customers personal services via personalized recommender systems. Recommender systems can provide customers information accurately, learns customer experience, and discovers knowledge. It also supports marketing strategy, increases customers’ loyalty and improves companies’ advantage.
Because there is numerous advantage of recommender systems, we propose a recommender system architecture on the basis of users’ browsing behavior. We use RFM model and self-organizing map to select target customers, analyze their browsing preference and discover the association rules between products. Finally we integrate customers’ browsing preference and association rules to provide recommend list. We attempt to forecast consumers’ behavior precisely to provide information they are interested in indeed, and assist companies in implementing marketing strategy more efficiently.
|
|---|
