Content-based personalised recommendation in virtual shopping environment

摘要

In our paper, we illustrate two kinds of product recommender algorithms to support e-commerce. For those commodities which a consumer seldom buys, user-rating methods are required to acquire the data set of the products rating in terms of the preference of the specific user. Thus, the combination of the Genetic Algorithm (GA) and k nearest neighbour method is proposed to infer the customer's personal preferences from rated products. On the other hand, for products that the consumers often buy, an interactive mode is provided for the users to evaluate the degree of interest for each feature of the products. We finally incorporate an intelligent agent model into the virtual shopping mall, which makes it easy for customers to fuse into the shopping experience.