"Detecting Deviators From the Norm - 'An Accurate Identification Method of Abnormal Users in Social Network Based on Multivariate Characteristics'"
Research published in the International Journal of Web Based Communities introduces a new method for identifying abnormal users in social networks, which involves analyzing multiple user behavior characteristics. Using the APIs of different social networks, Jian Xie of the College of Education at Fuyang Normal University in Fuyang, China, collected comprehensive data about users, including information about their accounts, the content they post, and the specific behaviors they exhibit. This data analysis allowed him to ascribe a set of attributes to users. Through attribute reduction, he eliminated redundant features and built a targeted attribute feature set to analyze suspicious accounts. Xie then used the data to train the XGBoost model, a Machine Learning (ML) algorithm, in order to develop a highly objective function that can quickly flag abnormal behavior on a social network. Xie was able to identify abnormal users with 95 percent accuracy. This level of accuracy in identification is enough to alert the system's administrators to any potential issues, which could then be manually investigated and handled (e.g., blocking malicious users). The approach could set the groundwork for developing highly effective social network security policies. This article continues to discuss the proposed approach to identifying abnormal users in social networks and its potential impact on security for social networking.