Research on Chinese Entity Relation Extraction Method Based on Deep Learning

摘要

With the continuous progress of deep learning, “Entity relation extraction” is the most important subtask of information extraction. The theory of entity relation extraction model based on deep learning in Chinese is mature gradually. The purpose of this paper is to introduce the concept, classification and characteristics of Chinese entity relation extraction and explain the evaluation index. Under the framework of deep learning extraction, aiming at the use of deep learning method in entity extraction, the main methods are divided into supervised method and distant supervision method. Based on the two methods, the derived methods are explained and compared in detail. Chinese domain relation extraction is due to different language structures and other reasons, and the applicability of the relation extraction model based on English is not good. Although it started late, with the efforts of researchers, the current method is relatively perfect. Finally, this paper predicts and prospects the development direction of Chinese entity relation extraction in the future.