A Possible Approach to Generalize NLP Models: Semantic Parsing with the “Term and Attribute” Structure
Abstract: With the attempt to convert all legal rules into structured knowledge, the author has manually analyzed more than 20,000 legal rules since 2008. Through the extensive and recurrent analysis and deducing, the author discovered that the essence of human languages is to describe the relationship among Persons, Acts and Objects. Based on this finding, the author developed the “Term and Attribute” structure, which can convert all languages into structured knowledge.
Paper Type: Short
Research Area: Special Theme (conference specific)
Research Area Keywords: Generalization of NLP Models, Semantic Parsing , Term and Attribute
Contribution Types: Theory
Languages Studied: Chinese, English
Submission Number: 8491
Loading