Survey on Event Extraction Technology in Information Extraction Research Area

Abstract

Event extraction is one of the most challenging tasks in the field of information extraction. The research of event extraction technology has important theoretical significance and wide application value. This paper makes a survey of event extraction technology, describes the tasks and related concepts of event extraction, analyzes, compares and generalizes the relevant descriptions in different fields. Then analyzes, compares and summarizes the three main methods of event extraction. These methods have their own advantages and disadvantages. The methods based on rules and templates are more mature, the method based on statistical machine learning is dominant, and the method based on deep learning is the future development trend. At the same time, this paper also reviews the research status and key technologies of event extraction, and finally summarizes the current challenges and future research trends of event extraction technology.