Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Knowledge graph generation method for online resource related information extraction

A technology of knowledge graph and related information, applied in the field of information extraction in natural language processing, can solve the problems of limited extraction accuracy, time-consuming and laborious online resource knowledge graph, insufficient description of online resource attribute description, etc., so as to improve construction efficiency and reduce The effect of labor costs

Active Publication Date: 2021-01-29
BEIJING INSTITUTE OF TECHNOLOGYGY
View PDF3 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The present invention aims to solve the problem of limited accuracy in extracting online resource-related information in scientific and technological literature by using existing entity and relationship extraction technologies, and the description of online resource attributes in related technologies is insufficient, and the online resource knowledge map is constructed manually Time-consuming and labor-intensive problems, a knowledge map generation method for online resource-related information extraction is proposed

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
  • Knowledge graph generation method for online resource related information extraction
  • Knowledge graph generation method for online resource related information extraction
  • Knowledge graph generation method for online resource related information extraction

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0041] First of all, it needs to be explained that the present invention is a knowledge map generation method for online resource related information extraction, the purpose of which is to extract fine-grained information related to resources in online resource reference context sentences in scientific and technological documents, and generate knowledge based on the extracted information Atlas. The fine-grained information extraction includes the extraction of scientific and technological entities related to online resources and the extraction of the relationship between target resource entities and scientific and technological entities related to online resources. The problem is formally defined as follows: given a context sentence s={w 1 ,w 2 ,...,w N}, such as the context sentence "We selected our vocabulary from terms (words and phrases) in WordNet lexicon", N is the sentence word sequence length (in the example given above, the context sentence is composed of 12 words, ...

Embodiment 2

[0098] (1) Experimental data setting

[0099] Before starting to introduce the experimental data, the online resource-related scientific and technological entities and the resource-entity relationship definitions that are the object of fine-grained information extraction in the present invention are firstly given.

[0100] For online resource-related scientific and technological entities related to target resource entities, there are six different entity categories, namely: task (Task), method (Method), data (Data), evaluation index (Metric) and general entity (Generic Term) ). Detailed definitions and examples of each technology entity category are explained below:

[0101] 1) Task (Task): Including the problem to be solved, the system to be built, the specific application scenario and other scientific and technological entity descriptions that can be completed as goals.

[0102] For example.: Language modeling, relation classification, transductive inference, tree parsing....

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 knowledge graph generation method for online resource related information extraction, and belongs to the technical field of natural language processing information extraction.The method comprises the following steps: enumerating an input online resource reference sentence to generate candidate spans, and learning token representation in the sentence based on a BERT encoder to obtain representation of each candidate span, so that two tasks of entity extraction and relationship extraction are converted into a classification calculation problem based on the span representation; weighting the target functions of the two tasks to obtain a joint target function, and performing joint training by using a multi-task learning strategy; a trained information extraction modelis applied to a large-scale scientific and technological literature corpus to generate a knowledge graph of online resources. According to the method, the problem of insufficient description of online resource attribute description by entity and relationship extraction is solved, the labor cost for constructing the online resource knowledge graph is reduced, and the knowledge graph generation efficiency is improved.

Description

technical field [0001] The invention relates to a method for generating a knowledge map for extracting related information of online resources, and relates to the technical field of information extraction in natural language processing. Background technique [0002] At present, the problem of metadata information extraction in scientific and technological literature has received more and more attention. However, in addition to common keywords, literature citations, scientific and technological entities, and entity relationships, online resources in scientific and technological literature are another important metadata. information, so far not received enough attention. [0003] With the continuous expansion of the scale of scientific and technological literature, the number of online resources cited in the literature is also increasing rapidly. How to discover, track and understand these online resources from the massive existing literature and the latest literature has beco...

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): G06F16/36
CPCG06F16/367
Inventor 冯冲赵赫唐雨馨
Owner BEIJING INSTITUTE OF TECHNOLOGYGY
Features
  • Generate Ideas
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com