Auto Service Recommendation Based On Improved Collaborative Filtering Algorithm And Its Application In Gas Station Sales Platform

With the rapid development of automobile aftermarket,automobile service has gradually become a necessity in daily life.How to provide the service of personalized recommendation basing on the personality information of users has become a research hotspot.Although both traditional and existing improved recommendation systems have alleviated the problems caused by “information overload” to a certain extent,the recommendation results cannot make accurate and personalized recommendations according to the information of owners due to the scenes of automobile service and its particular requirements of the users.Therefore,the algorithm used in the existing recommendation system needs to be improved to adapt to the special requirements of the automotive service.As an important part of automobile aftermarket,gas station is the main entrance and service station for users to choose automobile service.With the exposure of the problems of single sales model and the intensification of the industry competition in recent years,the internal management mode and external service mode of the traditional gas stations have been unable to meet the bilateral needs of gas station managers and vehicle owners.Especially under the requirement of the development of network and diversification of automobile service,the management and service mode of gas station need to be upgraded and improved.Firstly,in order to solve the traditional recommendation methods applied in the auto service recommended scenario,some improvements have been made respectively for the algorithm based on Pearson correlation coefficient of similarity calculation and prediction score parts,combining with the application of the auto service scenarios,according to the traditional collaborative filtering algorithm.Besides,some special attribute information has been included in the calculation of the weight to owners in parts of similarity calculation.And the product features of auto service also have been added to the end of score calculation.Through the simulation experiments and the comparisons with the recommendation results of traditional methods and existing improved methods,it is proved that the system with the improved algorithm can effectively improve the quality of automobile service recommendation and has better service recommendation effects.Secondly,on the basis of the previous part,the module of automobile service recommendation based on the improved collaborative filtering algorithm is further applied to the online sales service platform of gas stations.On the basis of in-depth analysis of the business process and technical feasibility of the gas station,the functions involved in the user and server of the system are divided in detail,and technologies such as Java language and MySQL database are adopted.The interactive interface of malls and its various functional modules are designed and developed based on the B/S framework and MVC development model.The main design function module is the auto service recommendation module which provides personalized service recommendation according to the user's personality information.The visual application of the automobile service recommendation module based on the improved collaborative filtering algorithm is realized in reality through the development of the system.