Multi-Task Identification of Entities, Relations, and Coreference for Scientific Knowledge Graph Construction
Многозадачная идентификация сущностей, отношений и кореференции для построения научного графа знаний
2018-01-01
SCID: 54.1/7m7kjdru
Discuss with AI
SCIERC datasetScientific Information Extractor (SCIIE)coreference resolutionmulti-task information extractionscientific knowledge graph construction
Figures from the paper
Abstract (AI)
We introduce a multi-task setup of identifying and classifying entities, relations, and coreference clusters in scientific articles.We create SCIERC, a dataset that includes annotations for all three tasks and develop a unified framework called Scientific Information Extractor (SCIIE) for with shared span representations.The multi-task setup reduces cascading errors between tasks and leverages cross-sentence relations through coreference links.Experiments show that our multi-task model outperforms previous models in scientific information extraction without using any domain-specific features.We further show that the framework supports construction of a scientific knowledge graph, which we use to analyze information in scientific literature. 1 Extracting nodes (entities) The SCIIE model extracts entities, their relations, and coreference
Key Findings
1
Joint training reduces cascading errors between extraction tasks and exploits cross-sentence relations through coreference links.
2
SCIIE provides a unified multi-task framework with shared span representations for identifying and classifying entities, relations, and coreference.
3
SCIIE supports constructing scientific knowledge graphs for analyzing information in scientific literature.
4
The multi-task model outperforms previous scientific information-extraction models without using domain-specific features.
5
The paper introduces SCIERC, a dataset jointly annotated for scientific entities, relations, and coreference clusters.
Research Object
scientific articles and their information for scientific knowledge graph construction
Research Subject
identification and classification of entities, relations, and coreference clusters, including their use in scientific knowledge graph construction
Publication Details
Publication Date
2018-01-01
Journal
Publisher
ISSN
Cited by
813
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest
References available in scid.ai1
Cited by5
A Frustratingly Easy Approach for Entity and Relation Extraction2021
Graph Neural Networks for Natural Language Processing: A Survey2023
Few-shot Named Entity Recognition: Definition, Taxonomy and Research Directions2023
Performance Optimization for Wireless Semantic Communications over Energy Harvesting Networks2022
Constructing Scholarly Knowledge Graphs From Sociology Research Papers: Incremental Changes in Cause-Effect Statements2025