Named Entity Recognition Based on Bi-LSTM and CRF-CEL

Abstract

Named entity recognition is a fundamental and important task in natural language processing in some specific fields, such as network security, medical diagnosis, law and so on. In these field, there are many complicated named entities with less contextual effective information and longer word length. Therefore, the goal of this study is to propose a model based on Bi-directional Long Short-term Memory Neural Network (Bi-LSTM) and Conditional Random Field (CRF), combined with Complex entity library (CEL), so as to improve the accuracy of complex entity recognition. The BiLSTM-CRF -CEL model that we constructed could achieve a relatively high F1 score as the number of complex entities increases, when the dimensionality is appropriate. It also achieved comparatively high performances among commonly used entity recognition models.