Dessign And Implementation Of Personalized Music Recommend System Based On Domain Ontology

With the fast development of digital music,more and more users choose digital music as consumer object,however,facing the overload music information on the Internet,users often fall into the music selection difficulties,so it is a problem worthy of further exploration in music domain that how to find the music that users interest from huge music quickly and accurately.To solve this problem,different music websites offer different solutions,including content based music recommendation system,user evaluation based music recommendation system,and user similarity based music recommendation system,etc.Although these solutions solve the needs of users partly,every recommendation system has shortcomings.Based on the background above,this paper investigates and studies many music websites and finds that different music website has different music classification criteria,which makes the whole music system become very large and complex.From the user's perspective,we hope to build a unified music classification system,so I construct the music domain ontology and form a music label system in this paper.On the basis of the music ontology,this paper proposes two hybrid recommendation algorithms,one is to take full account of users' information and behaviors,integrating of content-based and collaborative filtering recommendation algorithm and one is to fully consider the users' needs and preferences,integrating of content-based and rule-based recommendation algorithm.In this paper,I finally design and implement a personalized music recommend system based on domain ontology.This system can provide users with personalized music recommendation,music search,music data visualization and other functions.In this paper,I describe overall architecture of the system and implementations of each module in detail and complete system testing.