论文部分内容阅读
[目的/意义]更全面地理解用户标签使用行为的影响因素,为标签推荐和应用实践提供指导。[方法/过程]采用日志分析法,从人均标签个数、标签类型分布两个方面分析图书主题的影响,并以计算机、心理学、经济学、文学作品、绘画5类图书为例,基于社会化标注社区豆瓣网的用户数据进行实证分析。[结果/结论]图书主题会对用户的标签个数及类型分布具有显著影响:1在标签个数方面,计算机、心理学、经济学、文学作品主题图书的人均标签个数相近,显著高于绘画主题图书;2在标签类型分布上,根据5类图书间的差异可分为3组:计算机、心理学和经济学、文学作品和绘画。这一结论对图书标签推荐及图书分面检索和导航系统的设计具有参考作用。
[Purpose / Significance] A more complete understanding of the factors that influence the user’s tag usage behavior may provide guidance on how to recommend and apply the tag. [Method / Process] Using log analysis method, this paper analyzes the influence of the theme of the book from the aspects of the number of labels per person and the distribution of label types. Taking the five kinds of books of computer, psychology, economics, literature and painting as examples, Standardization of community Douban net user data for empirical analysis. [Results / Conclusion] The theme of the book will have a significant impact on the number of users and their type distribution: 1 In terms of the number of tags, the number of labeled topics per subject per subject in computer, psychology, economics and literary works is significantly higher than Painting theme books; 2 in the label type distribution, according to the differences between the five types of books can be divided into three groups: computer, psychology and economics, literature and painting. This conclusion is of reference to the recommendation of book labels and the design of book faceted search and navigation system.