基于DBLP的作者协作关系挖掘  

Co-authorship detecting based on DBLP

作　　者：赵彩虹[1] 杨艳[1] 

机构地区：[1]黑龙江大学计算机科学技术学院,哈尔滨150080

出　　处：《黑龙江大学自然科学学报》2010年第4期491-494,共4页Journal of Natural Science of Heilongjiang University

基　　金：黑龙江省教育厅科学技术研究面上项目(11551352)

摘　　要：对DBLP提供的XML数据文件的解析,得到协作者最多的作者和出版物最多的作者,得到DBLP上协助者和出版物数量都是按幂律尾分布的结论。对协作者300人以上的作者建立作者协作图,改进Betweenness快速算法在协作图上挖掘出中心作者。研究中心作者近3年发表的论文,得到其新的研究领域。实验证实得到了很好的结果,对指导计算机学者更好的把握学术前沿发展方向有一定的意义。Parsing the XML file based on DBLP,the authors were got the most coauthors who publish most papers.Find the conclusion that the number of coauthors and publications both followed the power law tail distribution.A co-authorship graph hard been areated for the authors who have more than 300 coauthors and detect the centric authors using improved betweenness fast algorithm.The new study fields were got,which the centric authors based on the papers they published in recent 3 years.The experiment shows it is a good result.It is significant to guide the computer scholars to hold the development direction of computer field.

关 键 词：DBLP 社会网络分析 作者协作关系 XML 

分 类 号：TP31[自动化与计算机技术—计算机软件与理论]