Relation extraction method and device based on transfer dependency relation and structure assistant

A technology of dependency and relationship extraction, which is applied in the fields of text retelling and relationship extraction, dependency syntax analysis, and named entity recognition. It can solve problems such as poor versatility, differences in dependency definitions and symbols, and high cost of manual labeling training corpus.

Active Publication Date: 2019-08-13
INSPUR QILU SOFTWARE IND
View PDF4 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0011] The technical task of the present invention is to provide a relation extraction method and device based on transitive dependencies and structural auxiliary words to solve how to deal with the extraction templates existing in the current relation extraction There are differences in the definition and symbols of the dependency relationship with the rule-dependent tree bank, and the versatility is poor, the selection of machine learning features is difficult, the cost of manual labeling training corpus is high, and the accuracy of automatic labeling training corpus is low, so as to improve the effect of relationship extraction.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Relation extraction method and device based on transfer dependency relation and structure assistant
  • Relation extraction method and device based on transfer dependency relation and structure assistant
  • Relation extraction method and device based on transfer dependency relation and structure assistant

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0165] as attached figure 1 As shown, the relationship extraction method based on transitive dependencies and structural auxiliary words of the present invention includes the following steps:

[0166] S1. Obtain named entities and dependencies: use natural language processing software to process sentences, and obtain named entities and dependencies between different components; as attached figure 2 As shown, the specific steps are as follows:

[0167] S101. Read the sentence to be processed;

[0168] S102, performing word segmentation on the sentence to be processed;

[0169] S103, marking the part of speech of the word segmentation vocabulary;

[0170] S104. Identifying the named entity A 1 、A 2 ,...,A n ;

[0171] S105, performing dependency syntax analysis on the sentence;

[0172] S106. Return the named entity and the dependency relationship.

[0173] Example: Take the sentence "The first president of race H in country X came to country G with a long history, and...

Embodiment 2

[0275] as attached Figure 13 As shown, the present invention's relation extraction device based on transitive dependencies and structural auxiliary words, the device includes,

[0276] Named entity and dependency acquisition unit M1 is used to obtain the named entity and dependency in the sentence; as attached Figure 14 As shown, the named entity and dependency acquisition unit M1 includes,

[0277] The sentence reading subunit M101 is used to read the sentences to be processed;

[0278] The word segmentation subunit M102 is used to separate the sentence to be processed into several words;

[0279] The part-of-speech tagging subunit M103 is used to tag the part-of-speech of each vocabulary in the sentence to be processed;

[0280] The named entity recognition subunit M104 is used to obtain the named entities in the sentence to be processed;

[0281] The dependency syntax analysis subunit M105 is used to obtain the semantic relationship between words in the sentence to be...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a relation extraction method and device based on a transfer dependency relation and a structure assistant, belonging to the field of relation extraction. The technical problemto be solved by the invention is how to process the dependency relationship definition and symbol difference of an extraction template and a rule dependency tree library existing in current relationship extraction and has poor universality, difficulty in machine learning feature selection, high cost of manually annotating training corpora and low accuracy of automatically annotating the training corpora. According to the technical scheme, the method comprises the following steps that S1, acquiring named entities and dependency relationships, wherein natural language processing software is usedfor processing sentences, and acquiring the dependency relationships between the named entities and different components; S2, extracting a relationship in the original sentence; and S3, simplifying the original sentences and extracting the relationship. The device comprises a named entity and a dependency relationship acquisition unit, an original sentence relationship triple extraction unit andan original sentence simplification and relationship triple extraction unit.

Description

technical field [0001] The invention relates to the fields of named entity recognition, dependency syntactic analysis, text retelling and relation extraction, in particular to a relation extraction method and device based on transitive dependency relations and structural auxiliary words. Background technique [0002] A relationship is a triple that describes the semantic connection between entities, and its form is (A, ɑ, B), where A and B are entities, and ɑ is the semantic connection between entities. A large number of relationships exist in natural texts. For example, the sentence "A city is the capital of country M" contains the relationship (city A, capital, country M), and the sentence "President B of country X came to country M" contains the relationship (country X , President, B) and (B, come, country M). Relational extraction is an important research content in the field of information extraction. It can establish dependency types between different entities, conver...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06F17/27
CPCG06F40/295
Inventor 王功明谢超郭运艳王建华
Owner INSPUR QILU SOFTWARE IND
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap