Mobile Commerce Application Personalized Recommendation Algorithm Research Based On The Pattern Of M&C

Now, Application（App） Store of mobile client products is growing,for people, allof the app is derived from a phrase "service" for the life. In the time of thematerialistic aspect, business is here and there, people also gradually improve therequirement for the quality of life. In the process of position, we need to findinformation, such as restaurants, department stores, hotels and entertainment venuesand other different interests. Price, environment, evaluation, product variety, distanceare able to the important factors that affect the user’s needs and interests. On themarket, has appeared more and more mobile business App, to satisfy the users in theuncertainty of the timing point. So this article mainly aimed at the mobile commerceapplication field,put forward the model and personalized recommendation algorithmto suitable for people to live now service requirements.This article mainly introduced the personalized recommendation algorithm basedon the pattern of Mobile service To Customer（M&C）mobile commerce application.Characteristics of the algorithm introduced the user interest model and collaborativefiltering with the concept of location of interest. For personalized service of mobileclient, location is the very important influence factors for user interest, so thetraditional two-dimensional matrix model improve Mobile application-User-Item(M-U-I)3D model. Mobile location in the scene information and application categoryset up a new dimension, make it more comprehensive contains the user in the processof mobile location situation information, also targeted push the user personalizedinformation service under different locations.Based on the personalized recommendation algorithm used Slope One algorithm,Slope One algorithm is introduced through the analysis of the characteristics. It avoidthe shortage of the algorithm, and add the position of interest. The concept consideredthe user interest similarity and location, and improvement into Location-Slope One（L-Slope One） algorithm. Similarity metric algorithm was improved, and narrowthe scope of the project collection. Through the calculation of project clusteringdecrease neighbor items, and reduced the time complexity and space complexity. And improved algorithm simulation verification through the experimental method.Experimental results show that the algorithm is suitable for the model based onmobile commerce applications. For the sensitivity of the position andrecommendations, algorithm was optimized and improved precision and betterresponse to the requirement of users in different location scenario. Provide users withhigh-quality personalized recommendation service and information.