Power term named entity identification method and device

Abstract

The embodiment of the invention discloses an electric power term named entity recognition method, which comprises the following steps: acquiring electric power text information, and inputting the electric power text information into a pre-trained word vector model to obtain a mapped text information vector with a preset dimension; inputting the text information vector into a pre-trained bidirectional long-short-term memory network model to obtain a bidirectional feature of the text information vector; and inputting the bidirectional features into a pre-trained conditional random field model, and outputting a mark sequence conforming to a semantic rule to realize entity identification of the power term. According to the invention, the conditional random field model is accessed behind the bidirectional long short-term memory network model; the problem of grammar errors caused by mutual independence of two-way long and short-term memory network model output is avoided, the conditional random field model can consider the sequence of marked text information output, it is guaranteed that output tags conform to normal grammar expression, and the recognition accuracy is guaranteed.

Description

technical field [0001] Embodiments of the present invention relate to the technical field of data mining, and in particular to a method and device for identifying named entity of electric power. Background technique [0002] Named Entity Recognition (NER) refers to a technology that identifies proper nouns or other specific terms in a specific field from text. Important basic work in application fields such as knowledge graphs, question answering systems, syntactic analysis, and machine translation. Entity recognition for text content can enable practitioners in related industries to quickly analyze and obtain information about specific entities from massive texts. [0003] At present, the methods to solve the problem of named entity recognition in various fields are mainly divided into the following three categories: [0004] (1) Rule-based methods, this type of method uses linguistic experts in related technical fields to manually construct rule templates, and the select...

AI Technical Summary

Problems solved by technology

[0009] To this end, the embodiments of the present invention provide a method and device for entity recognition of electric power terms to solve the problem of low recognition accuracy in the prior art due to unreasonable grammatical output and insufficient model generalization in traditional electric power text entity recognition

Images