Recommender System's Demand Characteristics Research Based On The Development Of Customer Relationship

As the relationship between corporation and clients develops, they have personalized interaction. In this process, the clients'demands for personalized recommendation system change constantly. The development is dynamic. This essay generalizes the classification of the development stage of client-corporation relationship in existing literature and puts forward three stages in the development of client-corporation relationship, i.e. Attraction stage, Build-up and maintenance stage and Fading stage. At different stages the clients' behavior is different and their responses to external stimulates are distinct. Therefore the study of personalized recommendation system should be based on in-depth analysis of clients'behavior at different stages and the research model should be constructed on the related theories in the field of client-corporation relationship.This essay is mainly about the influence of recommendation system on the clients' behavior at different stages of the development of client-corporation relationship. First, the personalized recommendation system characteristic research model, including technology factors and relationship factors, is constructed at the basis of TAM-trust theory and some empirical analyses are done about it with data collected. The result is that the clients at the attraction stage give response to the external stimulates through perceptive level. Meanwhile, the initial trust lays the foundation for the further development of the relationship. So the personalized recommendation system of this stage should be reasonable and feasible, able to eliminate the clients'concerns about privacy protection and security, provide the suggestion on recommendation system based on positive analysis of this stage.Second, in this essay the research model of personalized recommendation system at the build-up and maintenance stage is constructed on trust-commitment theory and verified with data collected. After the client and the corporation have actual interaction, emotional bond is formed and affective level response emerges through external stimulates. Then stable relationship is established and can influence the clients'behavior. At this stage, personalized recommendation system should focus on consolidating the clients'trust in corporation and increasing their satisfaction. At the same time, the corporation need increase the clients'perceptual transparency and controllability and provide accurate recommendation information, then provide algorithm optimization based on positive analysis.Last, in this essay the research model of personalized recommendation system at the fading stage is constructed on TRA, with empirical analysis. The characteristics of clients' behavior at the fading stage are explained through habit control factor and curiosity control factor. It is found that using the existing emotional bond and providing stimulates to reawaken the clients could influence their attitude and thus affect their intention of repurchase. The recommendation system of this stage should exhibit the novelty of the recommendation, provide high-quality unexpected recommendation and meanwhile increase intimacy with the clients, then provide algorithm optimization based on positive analysis.