User Behaviour Analytics Of Online Musical Media

With the mushrooming of mobile internet in this increasingly digital world,people prefer to listen music in music media more and more,such as Spotify,QQ music and Xiami music.And the study of users' behavior in online musical society attracts many interests.This thesis analyzes the user behavior on Xiami.com and the research conducted by this thesis can be divided into three parts.First,user characteristic quantity.By understanding the characteristics of user behaviors on the online musical media,this thesis selects fifteen characteristic quantities from six perspectives such as socioeconomics,demographics and sociality.And it studies the statistical distribution of these characteristic quantities and their correlations,provides user profile of Ximai.com in different aspects.Second,diversity of musical tastes.A person's musical preference can reveal a lot about their personality,and conversely,one's personality may play a role in one's diversity of musical tastes.In this thesis,the disparities between different genres are expressed,and the associations between the diversity and various variables of users are explored,by proposing a true diversity metric which effectively captures the diversity of users' musical tastes.Finally,vital users' identification.By mapping the “user-music” bipartite graph into two subgraphs,this thesis makes a discussion on the evaluation of the influence of node in the user network.Influential users in this network are identified,with the help of the H-index method on weighted network.Compared with the traditional indicator,coreness,this thesis shed light on the advantage of H-index method.All of these results analyze user behavior of online musical media and portray user profile in different aspects.They will play prior roles in the development of digital market for music.