An Entity Relationship Extraction Method Fused with Trigger Word Recognition Features

An entity relationship and feature recognition technology, applied in character and pattern recognition, special data processing applications, instruments, etc., can solve problems such as noise

Active Publication Date: 2021-05-04
BEIJING INSTITUTE OF TECHNOLOGYGY
View PDF8 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to solve the noise problem caused by the existing entity relationship extraction method that treats all words in the sentence equally, and proposes an entity relationship extraction method that integrates trigger word recognition features

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
  • An Entity Relationship Extraction Method Fused with Trigger Word Recognition Features
  • An Entity Relationship Extraction Method Fused with Trigger Word Recognition Features
  • An Entity Relationship Extraction Method Fused with Trigger Word Recognition Features

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0094] This embodiment illustrates the specific implementation of a method for extracting entity relationships that integrates trigger word recognition features according to the present invention.

[0095] figure 1 Shown is the flowchart of the method.

[0096] Step 1. Design a model to identify trigger words in sentences;

[0097] Step 1.1 marks the trigger words for the sentences in the data set, and for sentences with trigger words, for example, "In the Institute of Automation, Chinese Academy of Sciences, there is a Sino-French Joint Laboratory of Automation and Applied Mathematics"; the two entities of this sentence are "China Academy of Sciences Institute of Automation" and "Sino-French Joint Laboratory of Automation and Applied Mathematics", the entity relationship expressed is "ART / User-Owner-Inventor-Manufacturer". The word "有" in the sentence can directly express the entity relationship. Use braces to mark this word as a trigger word, which is used to reco...

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 relates to a method for extracting entity relations by integrating trigger word recognition features, and belongs to the technical field of information extraction in natural language processing. The method first marks the trigger word for the sentence in the data set; then uses the trigger word in the sentence, the two entities and the sentence type calculated according to the relative position of the two entities to calculate the label vector, and then calculates the label vector and the attention mechanism captured in the model. The relative entropy of sentence feature representation; then design an objective function of entity relationship extraction; optimize the relative entropy and the objective function of entity relationship extraction to train the model of entity relationship extraction. It not only solves the disadvantage that the existing entity relationship extraction method treats all words in the sentence equally, but also helps the entity relationship extraction model to extract features that are more useful for entity relationship classification; ACE2005 is 2.5% higher than the F1 score of the existing best entity relationship extraction method.

Description

technical field [0001] The invention relates to a method for extracting entity relations by integrating trigger word recognition features, and belongs to the technical field of information extraction in natural language processing. Background technique [0002] The task of entity relationship extraction is to give a sentence labeled with two entities and return the semantic relationship between the two entities. For example, "Yao Ming, under the influence of his father Yao Zhiyuan, also loves basketball very much", the two entities in the sentence are "Yao Ming" and "Yao Zhiyuan", and the relationship between the two entities is "father and son". [0003] Entity relationship extraction is an important supporting technology for information systems such as information retrieval and question answering systems. Entity relationship extraction transforms the output of information system from coarse-grained document level to fine-grained entity level. For example, in traditional ...

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 & AuthorityPatents(China)
IPC IPC(8): G06F40/295G06F40/284G06F16/35G06K9/62
CPCG06F16/35G06F18/253
Inventor辛欣王艳
OwnerBEIJING INSTITUTE OF TECHNOLOGYGY