基于深度学习的跨社交网络用户身份识别研究  

Research on Cross Social Network User Identification Based on Deep Learning

作　　者：栾孟孟 赵涛[1] 卞怡倩 LUAN Mengmeng;ZHAO Tao;BIAN Yiqian(School of Management Science and Engineering,Anhui University of Finance and Economics,Bengbu,Anhui 233030,China)

机构地区：[1]安徽财经大学管理科学与工程学院,安徽蚌埠233030

出　　处：《衡水学院学报》2022年第1期5-9,共5页Journal of Hengshui University

基　　金：安徽省教育厅自然科学研究项目(KJ2019A0656);安徽省自然科学研究基金项目(1608085QF145);安徽财经大学研究生科研创新基金项目(ACYC2020360)。

摘　　要：近年来跨社交网络用户身份识别技术成为众多学者研究的一个热点领域。但现有的跨社交网络身份识别技术存在诸如识别的成本高、准确率低和普适度不足等缺点。而依靠社交网络上的用户昵称采用深度学习方法进行身份识别可以克服这些缺点。该方法首先采用网络爬虫爬取了同一用户在社交网络Facebook和Twitter上的昵称对,然后在对数据清洗、转换、集成和处理后作为实验数据集;最后对实验数据集采用深度学习算法进行训练和识别。实验表明,基于深度学习的身份识别算法的准确率、精确率、召回率,和F1值分别达到了92.44%、94.29%、92.44%、93.11%和92.49%,结果优于传统机器学习算法以及该领域其他相似研究的识别效果,证实了该方法可以以较低数据获取成本实现较高识别效果并在不同社交网络使用。In recent years, with the gradual rise of social networks, cross-social network user identification technology has become a hot area of research by scholars, but the existing cross-social network identification technology has shortcomings such as high identification cost, low accuracy, and insufficient universality. However, relying on the nicknames of users on social networks, using deep learning method for identity recognition can overcome these shortcomings. Firstly, the network crawler technology is used to crawl the same user’s nickname pairs on Facebook and Twitter, then clean, transform, integrate and process the data;finally, the deep learning algorithm of the experimental data set is trained and identified. Experiments show that the accuracy, precision, recall and F1 value of our ID recognition algorithm using deep learning technology reach 92.44%, 94.29%, 92.44%, 93.11% and 92.49% respectively. The result is better than the recognition effect of traditional machine learning algorithms and other similar studies in this field, which proves that this method can achieve higher recognition effect with lower data acquisition cost and can be used in different social networks.

关 键 词：社交网络 用户昵称 深度学习 用户身份识别 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]