中文摘要：

为探查社会化标注中用户标注中标签使用主题特性,构建了社会化标签主题模型,定义了用户主题熵,从用户活跃度角度来刻画社会化标注中用户标签的主题鲜明性,结合Last.fm数据进行实证分析,结果表明：主题极鲜明和极模糊的用户较少,大部分用户主题鲜明程度介于两者之间;对于不活跃用户和极活跃用户,其标签主题随着活跃度的增加而增强,且增强速度相近;对于活跃用户,其标签主题鲜明性保持相对稳定。

英文摘要：

In order to explore the topic features of users＇ tags in social tagging,this paper constructs a topic model of social tagging and defines the topic entropy. The paper measures the topic distinctiveness of users＇ tags in social tagging from users＇ activity degree. Combined with the data from Last. fm,an empirical analysis is carried out. Results show that fewer users＇ topics are extremely distinctive or extremely vague,and most users＇ topics are in between. For inactive and highly active users,the topic distinctiveness is enhanced with the increase of activity degree,and the speed is same. For active users,the topic distinctiveness remains relatively stable.