A Survey on Deep Learning Event Extraction: Approaches and Applications

Area of Science:

• Natural Language Processing
• Artificial Intelligence
• Machine Learning

Background:

• Event extraction (EE) is vital for analyzing large text volumes.
• Deep learning has significantly advanced EE research.
• Existing literature lacks a comprehensive, updated survey of general domain EE.

Purpose of the Study:

• To provide a comprehensive survey of state-of-the-art deep learning-based event extraction methods.
• To classify general domain EE research based on task definitions.
• To identify future research directions in the field.

Main Methods:

• Literature review and classification of general domain EE.
• Detailed discussion of EE paradigms and deep learning models.
• Summary and comparison of benchmarks and evaluation metrics.

Main Results:

• A novel classification of current general domain EE research.
• Detailed overview of various EE approaches and their underlying models.
• Comprehensive comparison of different EE methods and their performance.

Conclusions:

• Deep learning-based EE is a rapidly evolving field.
• A structured understanding of current methods and benchmarks is crucial.
• Further research is needed to address emerging challenges and opportunities in EE.