Design And Implementation Of Multiple Rules Recommendation System Based On Location

With the rapid development of computer science,global Internet era has come.The development of Internet is gradually changing people's traditional way of life.In the era of the big data,information is exploding,people can receive a large amount of information from the network every day,the recommendation system with information overload come into being.With the rapid development of Internet,commerce personalized recommendation system become the focus of the research.In order to solve the problems of the overload of extracurricular resources and lack of real-time personalized recommendation of college students,this paper explores the accuracy and real time of the recommended system.This paper combines university management with recommend resource,and designs the campus resource recommendation system under the background of big data.Using complex event processing techniques to combine four different kinds of multidimensional,isomerism data source,real time location,check-in information,library lending and dormitory information,and used the EPL language to change simple events to complex events through the Esper engine,it implemented the whole process from student basic information data flow processing,complex event rule verification to related resource recommendation.Compared with collaborative filtering algorithm based on location,it is proved that the recommendation system performs well in real-time and accuracy.This paper mainly includes the following parts.(1)Discuss the design principle and the background of the campus extracurricular resources recommendation system based on students real-time location information,casual login in information,principle personal information,library information together to form a complex flow of events,and then put them into the Esper engine,matching the EPL rule to form the recommended results.(2)Introduce the key technologies in the recommender system,such as complex event processing,Esper engine,EPL language,collaborative filtering,hybrid KNN and Bias online positioning and so on.(3)Take advantage of the ability of real-time processing events of the powerful Esper engine,the recommending system in this paper recommends resources around the building to the students in real time according to the changing location information of students.Compared with the performance of recommender system based on collaborative filtering,the proposed recommendation system has been improved to performed well in the accuracy and real-time performance of recommendation.(4)Introduce the temporary interest and intimate friends set of these two recommended factors,taking the students' personal behavior and environment impact on their decisions combined into consider,exploring the effectiveness of the recommendation from multidimensional data sources.