Research on Named Entity Recognition of Elevator Safety Accident Domain
Knowledge map technology is an effective solution to solve the problem of multi-source heterogeneous data.It is ap-plied in many fields at present,and named entity recognition is a key step to automatically build the domain knowledge map.Howev-er,there is no related research on named entity recognition in the field of elevator safety accidents.Aiming at the application pur-pose of building the knowledge map of elevator safety accident field,this paper proposes a model based on the combination of BERT pre training model improved for Chinese text segmentation and BiLSTM-CRF to automatically extract entities from unstructured text in the field,and proposes a named entity recognition model suitable for elevator safety accident field.This paper collects and col-lates more than 500 elevator safety accident texts as the experimental corpus data set.Experiments show that compared with the tra-ditional named entity recognition model,the recognition effect of the model used in this paper is significantly improved.
国家科技期刊平台
NSTL
万方数据
