Group Recommendation Of Tea Products Based On Phrase LDA Model

With the development of network and information technology,the electronic commerce of agricultural products is developing rapidly.According to the national Ministry of commerce data display,2014 national agricultural e-commerce platform is booming,currently has exceeded 30000.With tea as the representative of the characteristics of agricultural products,but also to seize the opportunity,from the line to the line,through the network to thousands of households.Due to the explosive growth of agriculture e-commerce,online shopping population increased rapidly,how to find consistent with the user interest in the goods and how to a group have the same interests of users recommend commodities in the vast amounts of information become problems urgently to be solved.Recommendation system,the use of consumers in electric business platform of recessive or dominant behavior,analyze their preference,in the information overload,recommend to satisfy their interest preference.The current recommendation system is mainly aimed at individual,considering the sharp increase in the number of users and goods,one-on-one recommended cost is too high,how to use the recommended to group is a very popular research direction,recommend group not only can save the cost,the recommended and can be convenient user communication within the group,through group the recommendation trust familiar degree increase.Thus group recommendation system is helping agricultural electronic commerce development must use tools.Recommend the system,the use of consumers in the electronic business platform,the implicit or explicit behavior,analysis of its preferences,in the overload of information,recommended to meet their interest preferences of goods.The current recommendation system is mainly aimed at individuals,considering the rapid increase in the number of users and goods,one to one of the recommended cost is too high,how to use the group is a very popular research direction.This thesis introduces the related theory of existing group recommended,analyzes the tea under the environment of e-commerce recommendation problems and based on the shortcomings of the recommendation score,and puts forward the recommendation model based on Phrase-LDA model of tea group.The specific research contents are as follows:(1)Based on the comments of the tea products,the topic of the comments on the users and the tea products is extracted to express the preference of the users and the theme features of the tea products.The advantages and disadvantages of the traditional user preference representation method are analyzed,and the advantages of the other methods are analyzed.(2)In view of the traditional LDA is based on the word,the word relative phrase does not have the explicit meaning.This paper proposes Phrase-LDA,and uses it to integrate group user preferences.Phrase-LDA topic model on the one hand,the integration of the group user's preferences,on the other hand,the theme of the expression,the phrase of a clear semantic more detailed description of the user preferences.(3)Using the data of some tea in the Jingdong,the experiment was carried out to verify the validity of the proposed Phrase-LDA model,and a prototype system based on Phrase-LDA tea product was implemented.The research results of this thesis further refine the research of user preference.In addition,the method of this thesis has a certain novelty,which has a certain reference value.