Abstract: A modal dependency structure represents a web of connections in a document describing the source and epistemic strength of statements that helps to establish factuality in a given text.
Obtaining such graphs defines a core task of modal dependency parsing, which involves event and source identification as well as labeling of modal relations between them.
In this paper, we propose a simple yet effective biaffine modal dependency parser for English and Chinese that outperforms previous work.
Paper Type: Short
Research Area: Syntax: Tagging, Chunking and Parsing
Research Area Keywords: Information Extraction, Multilingualism and Cross-Lingual NLP, Syntax: Tagging, Chunking and Parsing
Contribution Types: NLP engineering experiment, Publicly available software and/or pre-trained models
Languages Studied: English, Chinese
Submission Number: 3159
Loading