With the development of science and technology and the prosperity of commerce,clothing e-commerce has entered a period of rapid development,and industry competition is becoming increasingly fierce,more and more clothing e-commerce platforms improve their core competitiveness through personalized recommendation.Based on the current understanding of personalized recommendation on clothing e-commerce platform,this paper analyzes the inadequacy of RFM model commonly used in clothing e-commerce platform,and adds two indicators of return rate and quantity of purchased goods on this basis to form RFMRQ model.At the same time,this model is combined with the heuristic collaborative filtering technology to build a personalized recommendation algorithm based on collaborative filtering of users and items,and carry out empirical analysis based on the historical data of clothing e-commerce platform.Through the practice test,it is found that RFMRQ model has more obvious improvement in the accuracy of user classification than the traditional RFM model.When constructing recommendation algorithm based on RFMRQ model,it is better to choose user-based collaborative filtering recommendation.
万方数据