图书情报工作 ›› 2017, Vol. 61 ›› Issue (6): 107-114.DOI: 10.13266/j.issn.0252-3116.2017.06.017

• 情报研究 • 上一篇    下一篇

利用实体解析的跨社交媒体同一用户识别

齐林峰   

  1. 上海大学图书情报档案系 上海 200444
  • 收稿日期:2016-12-14 修回日期:2017-02-20 出版日期:2017-03-20 发布日期:2017-03-20
  • 作者简介:齐林峰(ORCID:0000-0001-9809-4874),硕士研究生,E-mail:qlf317@gmail.com。

The Identity of the Same User with Cross-social Media Based on Entity Resolution

Qi Linfeng   

  1. Department of Library, Information and Archives, Shanghai University, Shanghai 200444
  • Received:2016-12-14 Revised:2017-02-20 Online:2017-03-20 Published:2017-03-20

摘要: [目的/意义] 跨领域关联实体一直是实体解析研究的主题,本文旨在不同的社交媒体(跨社交媒体)中找到属于同一用户的账户。[方法/过程] 在传统近似字符串匹配技术的基础上,提出使用属性值结合社交媒体中的链接和文本内容的方法,比较两个不同社交媒体账户的属性相似度、邻域相似度和关键词相似度这三个匹配函数,以此提高识别这两个账户是否是同一个人的精确度。并利用社交媒体Facebook和Twitter数据作为实验数据集,针对匹配函数的不同组合进行试验。[结果/结论] 结果表明,三个匹配函数的组合能够得到更多的账户匹配为同一用户,同时精确度也很高,达到0.923。本文提出的方法在Facebook和Twitter上的成功运用,给其他社交媒体平台或者其他领域的实体关联的研究提供了一条新的路径。

关键词: 社交媒体, 实体解析, 属性链接, 跨社交媒体

Abstract: [Purpose/significance] Associating entities across multiple domains has always been the subject of entity resolution, and the purpose of this paper is to find accounts that belong to the same person between different social media (cross-social media).[Method/process] Based on the traditional approximate string matching technique, the paper proposes the method of using attributes combined with links and text content in social media, and compares attribute similarity value, neighbor similarity and keyword similarity between the two different social media accounts, in order to improve the precision.[Result/conclusion] Using Facebook and Twitter as experimental datasets to test different combinations of matching function, the results show that the combination of three matching functions can get more accounts for the same user. At the same time, the precision is also high, and has reached 0.923. The successful application of the proposed method on Facebook and twitter provides a new path for the research of other social media platforms and other domains.

Key words: social media, entity resolution, attribute link, cross-social media

中图分类号: