Named entity extraction method and device

Abstract

An embodiment of the invention provides a named entity extraction method and device. The method includes: preprocessing a sample corpus to generate a standard corpus with the preset length and an annotation; constructing a feature vector of each character in the standard corpus; training a preset long-term and short-term memory network according to the feature vectors to obtain a predicted label value of each character; optimizing the preset conditional random field model according to the predicted label value and the true label value of each character; conducting named entity extraction usinga neural network model consisting of the long-term and short-term memory network and the conditional random field model. By means of the technical scheme, the feature vectors capable of characterizing more corpus features are constructed by controlling the length of the standard corpus. Further, by means of the method for optimizing the output sequence of the neural network model by using the conditional random field model, the named entity extraction accuracy is improved while the named entity extraction efficiency is guaranteed.

Description

technical field [0001] The present application relates to the technical field of natural language processing, in particular to a named entity extraction method and device. Background technique [0002] Named entities refer to the names of people, institutions, places, and all other entities that are identified by names. More broadly, named entities also include numbers, dates, currencies, addresses, and quantity phrases. The main task of named entity recognition (Named EntityRecognition NER) is to extract named entities from text and classify them. Named entity recognition is an essential part of many natural language processing technologies such as information extraction, information retrieval, machine translation, and question answering systems. [0003] The commonly used methods for named entity recognition mainly include: rule-based methods, statistics-based methods, and hybrid methods combining rules and statistics. Among them, the rule-based entity recognition method...

AI Technical Summary

Problems solved by technology

[0005] The embodiment of the present application provides a named entity extraction method and device to solve the problems of low accuracy and low efficiency of the named entity extraction method in the prior art

Images