Identification of Malicious Posts on the Dark Web Using Supervised Machine Learning

Abstract Given the constant growth and increasing sophistication of cyberattacks, cybersecurity can no longer rely solely on traditional defense techniques and tools. Proactive detection of cyber threats has become essential to help security teams identify potential risks and implement effective mitigation measures. Cyber Threat Intelligence (CTI) plays a key role by providing security analysts with evidence-based knowledge about cyber threats. CTI information can be extracted using various techniques and data sources; however, machine learning has proven promising. As for data sources, social networks and online discussion forums are commonly explored. In this study, we apply text mining techniques and supervised machine learning to data collected from Brazilian Portuguese Dark Web forums to identify malicious posts. Our contributions include the creation of three original datasets, including a labeled dataset with 26,575 posts, a novel multi-stage labeling process combining Indicators of Compromise (IoCs), contextual keywords, and manual analysis, and a comprehensive evaluation of text representations and classifiers. To the best of our knowledge, this is the first study to focus specifically on Brazilian Portuguese content in this domain. The best-performing model, based on LightGBM and TF-IDF Unigram, achieved 94% accuracy, 74% precision, 83% recall, and 79% F1-score. We also applied topic modeling to explore the model’s outputs on unlabeled data, observing thematic patterns consistent with those found in the labeled dataset. Overall, the results indicate that supervised machine learning combined with sparse text representations can effectively support the triage of CTI-relevant content in Brazilian Portuguese Dark Web forums, while also providing a lightweight and reproducible baseline for future studies in multilingual threat intelligence.