ABSTRACT
As is mentioned above, the analysis of characteristics of mobile terminals, for the recommendation of songs to users, can no longer be simply measured by the dominant information such as the user ratings, the past log information, but songs’ real listening time, repeat play count per unit time, and other factors such as the use condition of personalized service and the external information generated by the user in a particular situation should be considered to form a model of feedback and appraisal system in order to filter collaborative algorithms. [7] So, on the basis of predecessors’ research, this article will improve the traditional two dimensional model based on the usersextension project to a three dimensional model of the context of the user-projects-situation information. Among the information, the situation information consists of the user situation, the task situation and physical situation. The user information includes the basic information such as age,
1 INTRODUCTION
With the rapid development of mobile Internet technology and the increasing power of mobile intelligent terminal function, the wide use of mobile terminals can even be called a kind of dependence on a mobile terminal. Confusing music resources in Mobile music market has far exceeded a user’s music demand and capacity for independent choice. In a general situation, users usually, especially in a demanding moment, search for online music based on keyword search in the music platform. By measuring musical played record, ranking or word of mouth, user comments, they select the final network music that they decide to listen to. It is difficult to fully combine with rating and visit terminals such as shrimp music when searching a song. About this issue, the current solutions are usually based on the personalized information of users, such as downloading history, recommending the music that may cause their interests. Then they improve the network music’s “appearance” probability, which is regarded as one of the methods for precise delivery. However, one of the problems existing in the solution is the uncertainty of both the user’s access to the music platform and access to music.
