Li, Ying, et al. "NEDetector: Automatically extracting cybersecurity neologisms from hacker forums." Journal of Information Security and Applications 58 (2021): 102784.

The study introduces NEDetector, a novel method for identifying cybersecurity words and extracting neologisms from unstructured content on social media platforms, particularly hacker forums and Twitter. NEDetector uses a Bidirectional LSTM algorithm for identifying cybersecurity-related terms and a RandomForest algorithm for extracting new terms. The system achieved an identification precision of 89.11% and proved to be more timely than Google Trends in predicting cybersecurity threats from Twitter data.