Research And Application Of Agent-based Personalized Service Platform

Nowadays, it is the era of rapid development of the Internet, Internet technology is changing rapidly, and Internet applications come out one after the other. Questions that bother users have changed from the former insufficient resources to how to find resources that they are interested in from the complex and sufficient resources. To face the complex Internet resources, many users often feel powerless or even lost in the vast ocean of information. Therefore, the personalized service platform emerged, which is user-centered and provides services according to the user’s personality.Personalized recommendation system is an important research direction of personalized service. Based on the full study of Agent technology and personalized service, against the shortcomings of ignoring the user group sociality and user collaborative of current personalized systems, this thesis proposes a personalized service model based on Agent. This thesis designs an improved collaborative filtering algorithm based on user interest decay, this algorithm will be used in the personalized service model to improve the quality of personalized service. Learning Agent in the model uses the improved Apriori algorithm to analyze user behavior log to tap the user’s interest; Resource Agent uses the improved k-means algorithm to cluster resources; In Service Agent, a content-based and collaborative filtered adaptive recommended proportion of hybrid recommendation algorithm is proposed. Each Agent cooperates together to complete the goal of personalized service.Combined with today’s medical reform hotspots, the thesis designs and implements a case system of doctor-patient communication personalized service platform based on Agent, to prove the correctness and validity of the proposed model and algorithm through experimental analysis of the system.