A General Framework for Information Extraction using Dynamic Span Graphs

We introduce a general framework for several information extraction tasks that share span representations using dynamically constructed span graphs. The graphs are constructed by selecting the most confident entity spans and linking these nodes with confidence-weighted relation types and coreference...

Full description

Saved in:
Bibliographic Details
Published inarXiv.org
Main Authors Luan, Yi, Wadden, Dave, He, Luheng, Shah, Amy, Ostendorf, Mari, Hannaneh Hajishirzi
Format Paper Journal Article
LanguageEnglish
Published Ithaca Cornell University Library, arXiv.org 05.04.2019
Subjects
Computer Science - Computation and Language
Domains
Enumeration
Graphical representations
Graphs
Information retrieval
Online AccessGet full text

Cover

More Information
Summary:We introduce a general framework for several information extraction tasks that share span representations using dynamically constructed span graphs. The graphs are constructed by selecting the most confident entity spans and linking these nodes with confidence-weighted relation types and coreferences. The dynamic span graph allows coreference and relation type confidences to propagate through the graph to iteratively refine the span representations. This is unlike previous multi-task frameworks for information extraction in which the only interaction between tasks is in the shared first-layer LSTM. Our framework significantly outperforms the state-of-the-art on multiple information extraction tasks across multiple datasets reflecting different domains. We further observe that the span enumeration approach is good at detecting nested span entities, with significant F1 score improvement on the ACE dataset.
ISSN:2331-8422
DOI:10.48550/arxiv.1904.03296