Domain Knowledge Based Personalized Recommendation and Its Application
Yong Shi · Guanli xuebao · 2012
This paper discusses the traditional personalized recommendation methods and their applications in e-commerce.Then,it argues that classic collaborative filtering usually does not take domain knowledge into consideration and its limitation in real world application.It further discusses the concept of domain knowledge,and build a two-steps recommendation model which combines domain knowledge with traditional recommendation method.Finally,through empirical study,it is proved that personalized recommendation based on domain knowledge runs much better than traditional recommendation method in all evaluation indicators.