Cyberspace Event Extraction Techniques: From Traditional to Deep Learning Paradigms

Cyberspace event extraction is an important research filed of information extraction. Cyberspace event extraction converts unstructured data into structured formats by extracting event-critical information through various algorithms. This paper provides a comprehensive survey of event extraction methodologies, tracing their evolution from early rule-based systems to recent advances in large language model (LLM) based approaches. Firstly, the definition of event extraction is introduced. Secondly, the methods based on pattern matching and the methods based on machine learning are revisited. Thirdly, a comprehensive review of the current popular methods using deep learning are presented. Finally, the problems of event extraction are summarized, and the future research directions are discussed.