Internet-oriented Music Suggestion Based On Associated Rules Mining

With the rapid development of the Internet, information on the Internet grow explosively. It becomes more difficult for Internet users from access to information. The traditional search engine technology can not meet the needs of users. In this context, personalized suggestion system came into being.Music suggestion system is a vertical component of personalized suggestion system At this stage the domestic portals to provide music suggestion services are mainly Sina and Sohu, both of which provide the suggestion to single music search. We believe that the suggestion to single music search can not fully reflect the user's interest and requirement, coupled with the suggestion of the two results tend to suggest the same singers, what suggested that results lead to a lack of users' interest and requirements, so that the suggested results could not be satisfied. We need to know users' general interest of music on the basis of preference, to design a music suggestion system which users are really interested in and can explore potential requirements of the users. Most of personalized suggestions technologies based on the data mining, we reviewed data mining technology, inspired by some associated-rules mining and frequent flow mining algorithms, and put forward a well-established algorithms.. Subsequently we model a system on consideration of distributed system, then we design a system-wide distributed architecture and ultimately realize the music suggestion system.We test the user-satisfaction with the subjective test, as well as domestic similar music suggestion system Sina and Sohu. The results suggest that our music suggestion system was predominant.