Eureka AIR delivers breakthrough ideas for toughest innovation challenges, trusted by R&D personnel around the world.

Universal information extraction method and system based on unified structure generation

A general information and unified technology, applied in the direction of instrumentation, computing, semantic analysis, etc., can solve the problems of limited knowledge sharing, expensive and time-consuming data sets and knowledge sources, and achieve the effect of easy sharing

Active Publication Date: 2022-05-17
INST OF SOFTWARE - CHINESE ACAD OF SCI +1
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

First, developers have a lot of work to design and develop specific architectures for a large number of different information extraction tasks/settings/scenarios; second, learning isolated models for different data and

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
  • Universal information extraction method and system based on unified structure generation
  • Universal information extraction method and system based on unified structure generation
  • Universal information extraction method and system based on unified structure generation

Examples

Experimental program
Comparison scheme
Effect test

Example

[0085] (c) embodiments of the present invention will be extracted text and a specific structure extraction pattern guide input to the encoder - decoder architecture, directly generate a unified structured extraction language expression, and finally converted into a specific output structure. For example, for Test Example 1, embodiments of the present invention generate "((person: Steve(work for: Apple))"; for Test Example 2, embodiments of the present invention generate "(start position became(employee: Steve) (employer: Apple)... (organization:Apple))” For Test Example 3, embodiments of the present invention generate "((person: Steve) (organization: Apple) (time: 1997))". Finally, embodiments of the present invention is structurally transformed from the above expressions to generate a structured record after extraction.

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 provides a universal information extraction method and system based on unified structure generation, and belongs to the technical field of natural language processing, a universal structured extraction language is adopted to express different extraction structures, and the structured language comprises different hierarchies and can express information extraction results of various different structures; during decoding, a specific extraction demand is modeled through a structured framework extraction guide mechanism, and the model is helped to be quickly generalized to a specific task; the unified generation model is pre-trained by using different tasks, and the pre-trained model is finely adjusted, so that the performance of the unified generation model is improved.

Description

technical field [0001] The invention relates to a general information extraction method and system generated based on a unified structure, and belongs to the technical field of natural language processing. Background technique [0002] Universal Information Extraction aims to automatically extract structured information from unstructured text, such record information includes but not limited to text entity structure, relationship structure between entities, and multi-emotional structure. Taking entity relationship information extraction as an example, given the sentence "In 1997,Steve was excited to become the CEO of Apple.", an information extraction system should be able to identify an "employment" event whose trigger word is "become". The metastructure is "Steve" (subject), "Apple" (object) and "1997" (event); three entities, Apple: company, Steve: person, 1997: time; one relationship, "Steve" works for "Apple ". General information extraction is a key task in knowledge...

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
IPC IPC(8): G06F40/253G06F40/30G06K9/62
CPCG06F40/253G06F40/30G06F18/214
Inventor 孙乐陆垚杰韩先培林鸿宇肖欣延戴岱郑佳
Owner INST OF SOFTWARE - CHINESE ACAD OF SCI
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
Eureka Blog
Learn More
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com