Research And Application For Personalized Recommender System Of Cross-Border E-Commerce

With the vigorous development of Cross-Border-Commerce, the massive foreign goods are presented to people and localized in Electricity supplier warehouse, then it arises the phenomenon of information-overload. At the same time the customers feel difficult to select their favorite goods. The target consumer groups of Cross-Border-Commence is different from the traditional E-Commence’s consumer group. They pay more attention to the following aspects:securities of food、good quality、category diversity and reasonable price. The traditional personalize recommender system has already not met the need of Cross-Border-Commence. Under the context, the research of the Cross-Border-Commence-based recommender system is obviously urgency. It can help Cross-Border-Commence platform solve the problem of salement and relieve the inventory problem. Meanwhile it can select and recommend the consumer’s favorite item to consumer through the personalize recommender system.This paper firstly introduce the development situation of the Cross-Border-Commerce in our country, and then it has summarized and analyzed the context of the generation of recommender system% the current situation of research and several typical recommender systems algorithms. Based on the introduction of the main recommender algorithms, the paper is focus on introducing and analyzing the recommender system based on Collaborative Filter Algorithm. Against to the shortcoming of traditional multi-interest User-based algorithm, it presents an improved algorithm. The improved algorithm has considered the purchasing power of the comsumers、the good brand and so. The system practical data show that the improved algorithm has upgraded the rate of click and traffic conversion rate. It has effectively improved the customer satisfaction、the shopping experience of the customers. It also has solved the problem of inventory backlog.Secondly, it has realized the whole system of Cross-Border-Commerce which is based on personalized recommender, that includes the realization of the server the system framework and the client’s push framework.Lastly, it sums up the main research job of the paper and make a expect for the future research.