An Interpretable Gradient Boosting Framework for High-Fidelity Detection of Spambots and Artificial Followers in Social Networking Ecosystems
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
The emergence of social networking sites has further promoted the issue of spambots and counterfeiting followers, and these artificially inflate popularity rates, propagating false information, and losing the trust of the site user. The conventional methods of detection find it difficult to adapt to the changing spambot patterns and are not transparent in decision making. The paper proposes an explainable AI machine learning model at detecting spambots and fake followers with the help of gradient boosting tools, i.e., CatBoost, LightGBM, and XGBoost. The suggested approach combines profile attributes, behavioral patterns, and network-based characteristics to maximize the performance of detentions. Through the importance of the feature analysis, the interpretation of the model is clearly understood by establishing the factors that have influence in the predictions. It is experimentally proven that ensemble boosting models are better than baseline models in terms of precision and resiliency because both LightGBM and XGBoost show a higher classification rate. The results prove the hypothesis that high-performance machine learning combined with interpretability is an effective and reliable tool to fight malicious social network accounts.
Fields of Study & Subjects
Spam and Phishing Detection
Misinformation and Its Impacts
Hate Speech and Cyberbullying Detection
Publication details
- Journal / source
- Zenodo (CERN European Organization for Nuclear Research)
- Publication date
- 1 January 2026
- Language
- EN