Research On Personalized Recommendation Strategy In Mobile E-commerce

With the rapid development of mobile Internet,mobile communication technology,wireless sensor network technology,as well as the popularity of smart mobile devices,mobile e-commerce has penetrated into all areas of production,circulation and consumption of social economy and people's daily life,become a new direction for the development of electronic commerce.Compared with the traditional e-commerce,mobility and individuation are unique for mobile e-commerce.So personalized recommendation for modeling and optimization strategy has brought a series of problems need to be solved,the expected effect of the recommendation system should play,reflected in the following aspects:(1)the problem of real-time recommendation algorithm.(2)the problem that can not meet the needs of consumers and improve customer satisfaction.(3)the problem of enterprise sales and cross selling can not be improved.This paper uses the actual mobile e-commerce system as the application background.Based on previous studies,we analyzed and analyzed the characteristics of mobile e-commerce,and studied and designed a recommendation system model of mobile e-commerce based on the improved Apriori algorithm from the perspective of technology.In this paper,the improved Apriori algorithm is used to realize the unity of recommendation real time and accuracy.From the perspective of business application,the intelligent optimization method is used as a tool for modeling and model solving,so as to improve the consumer satisfaction and maximize the sales of the enterprise.Thus,it plays a positive role in promoting the faster and better development of the electronic commerce recommendation system in China.The specific research work and innovation of this paper are as follows:(1)Improvement and application of mining algorithm for mobile electronic recommendation system.Improvement and application of mining algorithm through the analysis of the present development of mobile e-commerce,mobile e-commerce recommendation system has advantages and defects,the traditional e-commerce recommendation system using Apriori algorithm needs to repeatedly scan the database,frequent item mining efficiency is low,this paper proposes an improved Apriorialgorithm.Mining for mobile e-commerce transaction data transaction data in the database,unify the real-time and accurate recommendation,finally to taobao.com a shop transaction data as the object,the detailed operation method and experimental results.(2)A multi-source information fusion recommendation strategy based on the consumer's initiative decision.A multi-source information fusion method and a recommendation product optimization method based on the consumer's initiative decision are put forward.In this paper,we propose to collect consumer information from third sources of smart phones from multi-source channels in mobile e-commerce,and then use the improved radial basis function(RBF)neural network to set weights.On this basis,the improved Dempster-Shafer(D-S)evidence theory is used to carry out information fusion and power spectrum estimation for product recommendation.We can effectively retain and make use of consumer behavior and consumption data in different information sources,and combine consumer satisfaction prediction method based on online active decision-making of consumers,recommend consumers' real satisfaction products,and optimize personalized recommendation results.Finally,the proposed classification model is verified on a real dataset.The experimental results show that the proposed method can better provide personalized recommendation for mobile users compared with other traditional methods.(3)Modeling and optimization of personalized freight strategy oriented to recommended products.The proposed personalized freight discount model to attract consumers to buy recommended products,to the enterprise of a product for shipping discounts as the background,the use of promotional product discounts to attract the attention of consumers,and then recommend related products to meet consumer preferences.In this paper,a comprehensive optimization model of freight discount and recommended product set is given,and a model solving method based on genetic algorithm is proposed.The proposed model can not only help enterprises make the best freight discount,but also avoid the loss of freight discount while attracting customers' traffic,and also can make the best product recommendation combination according to the freight discount and maximize the sales promotion benefits of enterprises.Finally,a case study is used to illustrate the effectiveness of the model,and sensitivity analysis is carried out to verify that the model can handle various kinds of products and changing consumer demand,enabling the electronic retailer to make better decision behavior.This paper studies the problems faced by the mobile e-commercerecommendation system,which has been solved from the above three aspects.To realize the real-time recommendation system by improved Apriori algorithm;through the proposed optimization model and solving the model construction method of multi-source information fusion model based on active consumer decision,effectively improve the customer satisfaction;by constructing the freight discount strategy model for recommended products,effectively solve the problem of enterprise sales and cross selling.Therefore,this research results promote the development of related research in personalized recommendation field to some extent,expand the idea of personalized marketing strategy of mobile e-commerce,and provide favorable theoretical and method support for practical application.