摘要
在信息环境下,关于网络信息行为的研究是一个有着普遍需求和意义的重要课题。高校大学生群体是中国网民群体中重要的核心力量和生力军,探索与把握当前中国大学生用户网络行为和兴趣需求特征具有丰富的现实意义与社会价值。可视化分析可以直观地呈现用户行为的整体分布特征,为进一步的深入分析奠定基础。本文以中国大学生的纵向网络访问日志为分析对象,通过实证分析揭示了大学生网络行为在学期、周、小时等多重时间粒度下的分布特征。同时,基于马尔可夫链、Gini指数、h指数等指标,本文进一步讨论了不同小时时段下大学生用户群体的行为时序性行为特征和兴趣需求分布,为大数据环境下理解大学生的网络生活规律和支持企业个性化推荐服务提供了科学参考。特别地,本文创新地将h指数应用于用户兴趣网站排名算法中,展现了经典信息计量学分析方法在网络用户行为分析过程与应用中的价值,促进了应用情报学不同方法之间融合。
In the information dependent environment,research on web behaviors is an important topic with widespread needs,especially concerning the analysis and research of online behaviors of college students.College students are an important core group and a new force among Chinese netizens.There is great practical significance and social value to exploring and grasping Chinese college students?characters such as user behavior,interests,and needs.Visual analysis can directly display the overall distribution characteristics of user behavior and lay the foundation for further in-depth analyses.In this study,the college students?network access logs are considered as the analysis object to indicate the features of group behavior under multiple time granularities,namely,term,week,and hour.Meanwhile,on the basis of research using Markov chain,Gini-index,H-index,and other feature indicators,this work attempts to reveal college students?online characters of sequential behavior,user interest,and needs in various hour intervals,which provides a scientific reference to understand the nuances of college students'online life and support enterprise personalized services under the big data environment.In particular,H-index is applied to the website ranking algorithm of user interest and shows the value of classical informetrics analysis in the process of analysis and application of online user behavior,thus promoting the integration of different methods in applied information science.
作者
严承希
王军
Yan Chengxi;Wang Jun(Department of Information Management,Peking University,Beijing 100871)
出处
《情报学报》
CSSCI
CSCD
北大核心
2018年第9期890-904,共15页
Journal of the China Society for Scientific and Technical Information
