Named entity identification method and apparatus, and terminal device

Abstract

The invention provides a named entity identification method and apparatus, and a terminal device. The named entity identification method comprises the steps of obtaining to-be-identified data; preprocessing the to-be-identified data to obtain a character vector of the to-be-identified data; inputting the character vector into a pre-trained character-based named entity recognition model, and outputting a recognition result representing named entity information of to-be-identified data; determining a named entity of the to-be-identified data according to an identification result; wherein the training feature vector of the named entity recognition model comprises an entity boundary feature vector of training data, and when the training feature vector is used for extracting features of the training data, normalizing a plurality of character features of the same named entity into features of corresponding named entities so as to perform model parameter training based on the normalized features of the named entities. According to the embodiment of the invention, the entity boundary problem existing in an existing character-based named entity recognition model can be solved, so that the model performance is improved, and the named entity recognition accuracy is improved.

Description

technical field [0001] The present invention relates to the technical field of data processing, in particular to a named entity recognition method, device and terminal equipment. Background technique [0002] Named Entity Recognition (NER) refers to the identification of entities with specific meaning in text or strings, mainly including names of people, places, institutions and proper nouns. Judging whether a named entity is correctly identified mainly includes two aspects, namely: whether the boundary of the entity is correct, and whether the type of the entity is marked correctly. [0003] Existing named entity recognition methods mainly include: rule-based methods and statistics-based methods. In this statistics-based method, the network structure commonly used at present is BI-LSTM-CRF (Bi-directional Long Short-Term Memory-Conditional Random Field, bidirectional long-term short-term memory neural network and conditional random field), the BI-LSTM- CRF can combine BI-...

AI Technical Summary

Problems solved by technology

[0005] Embodiments of the present invention provide a named entity recognition method, device, and terminal equipment to solve the problem of low accuracy in named entity recognition due to the existence of entity boundaries in existing character-based named entity recognition models

Images