Full-Text Search:
Home|Journal Papers|About CNKI|User Service|FAQ|Contact Us|中文
《Sciencepaper Online》 2011-01
Add to Favorite Get Latest Update

Probabilistic model for academic social network and its applications

Tang Jie1,Gong Jibing2,Liu liu1,Yang Wenjun3(1.Department of Computer Science and Technology,Tsinghua University,Beijing 100084,China;2.Department of Computer Science and Engineer,Yanshan University,Qinhuangdao,Hebei 066004,China;3.Information Center,Planning and Engineering Institute Petrochina Corp.Ltd,Beijing 100083,China)  
Previously,the different type of information has been investigated mainly in a separated way,which ignores the inter-dependencies.In this paper,we focus on:1) how to simultaneously model the heterogeneous data in an academic network;2) how to take advantage of networking information to improve the modeling effectiveness.We propose a novel author-conference citation topic model,named ACCT,to simultaneously model the topical aspects of the heterogeneous data(e.g.papers,authors,conferences and citations) in the academic network.The probabilistic model captures the semantic dependencies between different types of information with a latent topic layer.The modeling results can be used for both academic search and academic suggestion.Experiments on the ArnetMiner data show that the proposed model outperforms other baseline models.
【Fund】: 高等学校博士学科点专项科研基金资助项目(20070003093);; 国家高技术研究发展计划(863计划)资助项目(2009AA01Z138)
【CateGory Index】: O157.5
Download(CAJ format) Download(PDF format)
CAJViewer7.0 supports all the CNKI file formats; AdobeReader only supports the PDF format.
©2006 Tsinghua Tongfang Knowledge Network Technology Co., Ltd.(Beijing)(TTKN) All rights reserved