Design & Implementation Of Group Buying Recommendation System Based On Analysis Of Composable Consumer Behavior

With the development of Internet and information technology, "online group buying " this new e-commerce model has developed rapidly, and it attractes large businesses and consumers. The group buying website merchandise display mode still exists the problem of information overload. Consumers quickly find their interested products is still a difficult thing. With the rapid development in the field of group buying, how to dig the potential demand of users and provide recommendation service better for users has become an important research topic.The traditional recommendation algorithm using collaborative filtering method combined with user preferences realize personalized recommendation. But the group buying mode has several differences points with the traditional electronic commerce. Its consumer behavior is impacted by time, location, date and other factors, so the traditional recommendation algorithm is not well suited for group buying system. This paper analyzes different point between group buying mode and traditional electronic commerce, based on the analysis of combination consumer behavior of users, combined with association rules mining and logistic regression model etc., into the personalized user preference information, it realized the group buying website recommendation system. It has established a comprehensive and effective evaluation system and verifys the effectiveness of the recommend method is it is through the hit rate, On Line CTR, On Line ROI and other indicators in the contrast experiment.Group buying recommendation system proposed in this paper is mainly composed of two parts: off-line mining and online services. The offline part mainly through the mining association rules algorithm of user consumption data mining association and the off-line training of logistic regression model. Online services are mainly through the logistic regression model to sort the results of off-line recommendation and combined with the combination of users of consumer behavior analysis data to filter and adjust rules right.This paper first introduces the research background and research status at home and abroad, and explains the main research contents of this paper; then intruduces the project needs analysis. Then it explains the recommendation system evaluation system he recommendation system evaluation system including the evaluation method and evaluation index. The core of this paper is to study the combination of users of consumer behavior. It main includes the relationship between consumer behavior and single category, the relationship between consumer behavior and the distance of shops, the relationship between consumer behavior and time,date, permanent point; the following is part of the design and implementation of recommendation system. Finally, to assess the effect of recommendation system, and make the corresponding evaluation...