Design Of Presale Recommendation System For Automobile Service-oriented Lifecycle

The rapid development of the automobile industry is an indisputable fact,competition and consumers’changing demands have prompted dealers and manufacturers to provide better service:grasping the accurate market demands,innovating sales service mode and developing the vehicle models which cater to consumers’demands.With the development of information technology, obtaining information from the network has become a habit.Then establishing a automobile presale recommendation system based on Internet which can help customers look for information and innovate service mode and grasp the market quotations and further improve sales is a feasible scheme.The goal of this thesis is to establish a automobile presale recommendation system for service-oriented lifecycle which provides customers with presale personalized recommendation and decision support to dealers and manufacturers. The main research content is divided into the following several aspects:First of all,on the background that automobile service industry has brought high profits and it develops in low level in China,this thesis leads to the topic of presale recommendation service.On this basis,this thesis discusses the realistic requirements of customers, dealers and manufacturers:personalized recommendation, decision support and management of the potential customers. This thesis designs a automobile presale recommendation system supported by recommendation module and potential customers management module in view of the demand analysis.Secondly, this thesis discusses the relevant data with the center of customers which is used on the basis of demand analysis,and it uses SSIS (SQL Server Integration Services) platform to establish the data warehouse on the subject of customer analysis which is the foundation for subsequent recommendations.Once again, in view of the recommendation module,this thesis analyzes and compares the current mainstream recommendation technologies,and adopts the user-based collaborative filtering recommendation method. As this thesis involves large amount of data which affects real-time recommendation,this thesis adds off-line user cluster based on collaborative filtering.Potential customer management module includes functions as bellow:potential customers registration and management,potential customers tracking and management,negotiations registration and management and statistics of customers who call or come to store.Finally, this thesis uses four layers B/S structure to develop the whole system, and it realizes the recommendation module and potential customers management module,and at the same time it also validates the feasibility of the design.