Research Article
Across Online Social Network User Identification Based on Usernames
@INPROCEEDINGS{10.1007/978-3-031-30237-4_11, author={Zijian Li and Di Lin and Peidong Li}, title={Across Online Social Network User Identification Based on Usernames}, proceedings={Machine Learning and Intelligent Communication. 7th EAI International Conference, MLICOM 2022, Virtual Event, October 23-24, 2022, Proceedings}, proceedings_a={MLICOM}, year={2023}, month={4}, keywords={Across Social Network Similarity Feature Extraction}, doi={10.1007/978-3-031-30237-4_11} }- Zijian Li
Di Lin
Peidong Li
Year: 2023
Across Online Social Network User Identification Based on Usernames
MLICOM
Springer
DOI: 10.1007/978-3-031-30237-4_11
Abstract
Cross social network user identification aims to identify the same entity on various online social networks to enhance the completeness and accuracy of the persona. There are three broad categories of cross-social network user identification methods: user identification on account of basic user information, user identification on the basis of network topology graphs, and user identification based on the user's origin. This paper analyzes users’ display names from different social networks to determine whether they are the same person. The process consists of three steps: first, we obtain information about users and bring their display names from social networking sites. Secondly, we analyze the user's name, get a series of values from the user's name through similarity calculation methods, and match the similarity. We perform similarity matching on the real dataset by using some classification models. Our model performs well, with F1 values reaching 97.07%, 94.65%, and 92.05% for the three datasets, respectively. This paper can provide a high-quality dataset for downstream NLP tasks of high research significance and value.
