Abstract :

Social networking systems such as X (Twitter) function as centers for open human interaction; nonetheless, they are progressively permeated by automated accounts impersonating real users. These bots often participate in disseminating misinformation and influencing public sentiment at politically critical periods, such as elections. Many contemporary bot detection techniques depend on black-box algorithms, which raises issues about their transparency and practical applicability. This work seeks to overcome these constraints by formulating an innovative way for identifying spambots and counterfeit followers via annotated data.We present an interpretable machine learning (ML) framework that utilizes various ML algorithms with hyperparameters tuned by cross-validation to improve the detection process.Additionally, we examine several attributes and provide a distinctive feature set targeted for superior performance in bot identification.Furthermore, we use many interpretable AI methodol