电信科学 ›› 2015, Vol. 31 ›› Issue (7): 43-51.doi: 10.11959/j.issn.1000-0801.2015166

• 研究与开发 • 上一篇    下一篇

异质信息网络分析及其语义探索

王锐,张志强,石川   

  1. 北京邮电大学智能通信软件与多媒体北京市重点实验室 北京 100876
  • 出版日期:2015-08-21 发布日期:2015-08-21
  • 基金资助:
    国家重点基础研究发展计划(“973”计划)基金资助项目;国家高技术研究发展计划(“863”计划)基金资助项目;国家高技术研究发展计划(“863”计划)基金资助项目;国家自然科学基金资助项目;国家自然科学基金资助项目

Analysis and Semantic Mining in Heterogeneous Information Network

Rui Wang,Zhiqiang Zhang,Chuan Shi   

  1. Beijing Key Lab of Intelligent Telecommunication Software and Multimedia,Beijing University of Posts and Telecommunications,Beijing 100876,China
  • Online:2015-08-21 Published:2015-08-21
  • Supported by:
    The National Basic Research Program of China(973 Program);The National High Technology Research and Development Program of China(863 Program);The National High Technology Research and Development Program of China(863 Program);The National Natural Science Foundation of China;The National Natural Science Foundation of China

摘要:

由多种类型的实体和关系构成的异质信息网络逐渐成为社会网络分析的研究热点。作为异质信息网络的一个独特属性,元路径包含了丰富的语义信息。实际生活中的许多网络都包含带权值的链接,这使得不考虑链接上权值的传统元路径不能精确地捕捉网络中的语义信息。基于此,描述了异质信息网络的相关概念,并对异质信息网络的应用进行了简要介绍。通过将传统元路径扩展为带权元路径,更精确地描述了带权值的异质信息网络中微妙的语义信息。通过在两个真实数据集上进行实验,说明了带权元路径在推荐、相关性搜索中的应用效果。

关键词: 异质信息网络, 元路径, 相关性搜索, 推荐

Abstract:

Heterogeneous information network(HIN),which is composed of different types of objects and links,has gradually become a hot topic in social network analysis.As a unique characteristic of HIN,meta path contains rich semantic information.Heterogeneous information network with values on links are ubiquitous in real world.Therefore,the traditional meta path,which doesn’t consider weight on links,can not exactly capture semantics in many cases.Related concepts of HIN were introduced and a brief introduction of applications of HIN was given.Then subtle semantic information in HIN was explored by extending the traditional meta path to weighted meta path.Experiments on two real data sets demonstrate the applications of the weighted meta path in recommendation,relevance search.

Key words: heterogeneous information network, meta path, relevance search, recommendation

No Suggested Reading articles found!