| 引用本文： |
-
薛晖,孙波,司成祥,张伟,房婧.跨社交网络用户身份链接回顾与展望[J].信息安全学报,已采用 [点击复制]
- xuehui,sunbo,sichengxiang,zhangwei,fangjing.Advance in user identity linkage across online social networks[J].Journal of Cyber Security,Accept [点击复制]
|
|
| 摘要: |
| 随着互联网的飞速发展，社交网络平台（又称在线社交网络）也日益普及和多样化，为了更好地利用每个社交网络平台提供的服务，用户往往会加入多个社交网络平台。链接同一个自然人在不同社交网络平台中的账户，称为用户身份链接。通过用户身份链接可以充分了解用户的兴趣，极大的丰富用户画像，进而用于数字营销和推荐系统。本文回顾了用户身份链接方法的发展历史，根据其特征提取和模型构建两个阶段对现有用户身份链接方法进行了分类分析，讨论了存在的问题与挑战，总结了现有方法所采用的数据集和评估指标，最后展望了用户身份链接的未来研究趋势。本文通过提出一种用户身份链接问题的通用定义、比较分析已有用户身份链接方法、讨论存在的问题和展望未来研究趋势，将用户身份链接问题的现状和未来以清晰的结构化的方式进行分析展示，有助于研究人员对该领域的相关研究形成系统性的理解和把握，进而做出更加深入的研究工作。 |
| 关键词: 社交网络 用户身份链接 账号关联 锚链接预测 用户画像 |
| DOI：10.19363/J.cnki.cn10-1380/tn.2023.08.06 |
| 投稿时间：2020-12-30修订日期：2021-03-15 |
| 基金项目: |
|
| Advance in user identity linkage across online social networks |
|
xuehui1, sunbo2, sichengxiang2, zhangwei2, fangjing2
|
| (1.IIE, CAS;2.CN/CERT) |
| Abstract: |
| Nowadays, online social networks(OSNs) have become popular and diverse. To make better use of the services offered by each OSN, people tend to join a number of OSNs. User identity linkage across OSNs is the task of accurately link accounts corresponding to the same natural person in multiple OSNs. We can deeply understand user interests and greatly enrich user profile by user identity linkage, or use it in digital marketing and recommendation system. In this paper, we classify and analyze the existing user identity linkage methods according to feature type, and discuss existing problems and chal-lenges. We also summarize the datasets and evaluation indicators used by existing methods, then analyze why public agreed datasets are rare. Finally we look forward to the future research trend of user identity linkage. This paper proposes a general definition of user identity linkage, compares and analyzes existing methods of user identity linkage, discusses existing problems and looks forward to future research trend, to analyze and display current situation and future of user identity linkage in a clear and structured way, which is helpful for researchers to form a systematic understanding and grasp of the related research in this domain, so as to further make more in-depth research work. |
| Key words: online social network user identity linkage account association anchor link prediction user profile |
