Research On Retail Commodity Recommendation System Based On Collaborative Filtering Algorithm In B Company

In recent years,E-commerce companies have experienced rapid development.Now they begin to face many real problems and challenges such as the slowdown in the number of online users,the gradual shrinkage of the traffic dividends.Meanwhile,with the development of our country's economy and the improvement of residents' income levels,people pay more and more attention to offline shopping experience.The shortcoming of online e-commerce in the aspect of meeting user's shopping experience is becoming more and more prominent.Company B is a representative company in the new retail field and provides personalized services to users,and its main line of business is chain convenience store.So building a retail commodity recommendation system that meets its own characteristics is of great significance to B.In this article,considering the area of the convenience stores in the B company is small,and the goods sold in those convenience stores are mainly the hot commodities that can meet the immediate needs of the nearby users,the user based collaborative filtering algorithm is selected as the main algorithm for the construction of the retail commodity recommendation system of B company.In this paper,the algorithm is improved in two aspects.Firstly,aiming at solving the problem of cold start of collaborative filtering algorithm,this paper considers the user characteristic factors into the process of constructing the commodity recommendation algorithm model.Secondly,considering retail commodity recommendation system in B has a high requirement for the recommendation information's timeliness,this article take the time scene factor into consideration in the process of building the commodity recommendation algorithm model.By this way,the accuracy of the recommendation system can be improved.Finally,based on B's retail commodity recommendation algorithm model,this article conducts a demand analysis,design and implementation of B's retail commodity recommendation system,and tests the developed system.The results show that the retail commodity recommendation system which implemented in this paper meet the corresponding functional requirements.This paper constructs a retail commodity recommendation system which meets the operating characteristics of B company.This system can increase the efficiency of the user's shopping in the offline convenience store of company B and improve the user's shopping experience.Meanwhile,this research can also provide some references for other new retail companies which own offline stores to build their own commodity recommendation systems.