Implementation On The Crucial Techniques Of Music Recommend System

In this paper we study a new type of music recommendation,based on the combination of location-based music recommendation in mobile environment and the influence of the community to the accuracy of the recommendation.Different from traditional music recommendation systems,we will take user's current geographical ambiene into consideration in situational music recommendation first.Mobile user's location is constantly changing as well as his geographical ambience.User's feeling and expectation of music is significantly influenced by his inner-state which in turn is greatly affected by his geographical ambience,thus we have to make different recommendations according to user's geographical ambience.At the same time,this recommendation system considers the influence of the community.The algorithm analysis kinds of characteristics of music and user feelings to analysis users' behaviors.In this paper we propose a new music recommendation prototype named Music Network.Music utilizes location-based image retrival and other Web services to recommend music for users virtually located at any position.Then this web use the influence of the community to improve the accurancy of recommendation.We analysis the architecture and recommendation process of Music in depth.We tackle mainly three difficult problems in this paper.The first is arbitrary English word similarity calculation,the effectiveness and efficiency of word similarity calculation is very important to Music.Then we analysis the characteristics of music and users' feelings and combined with the influence of the communicity to improve the accurancy.The last is how to integrate the two algorithms.At the end of the paper,we show the Music to make reader know this system better.