Commonsense Knowledge Mining from Term Definitions

TL;DR

This paper explores extracting commonsense knowledge triples from dictionary term definitions using machine learning, demonstrating some valid and novel triples but also highlighting challenges with current scoring models.

Contribution

It introduces a method to mine commonsense triples from definitions and evaluates existing models, revealing their limitations and potential for capturing new knowledge.

Findings

Definitions contain valid and novel triples for some relations

Existing triple scoring models face challenges with this data

Initial evaluation shows promise but needs improvement

Abstract

Commonsense knowledge has proven to be beneficial to a variety of application areas, including question answering and natural language understanding. Previous work explored collecting commonsense knowledge triples automatically from text to increase the coverage of current commonsense knowledge graphs. We investigate a few machine learning approaches to mining commonsense knowledge triples using dictionary term definitions as inputs and provide some initial evaluation of the results. We start from extracting candidate triples using part-of-speech tag patterns from text, and then compare the performance of three existing models for triple scoring. Our experiments show that term definitions contain some valid and novel commonsense knowledge triples for some semantic relations, and also indicate some challenges with using existing triple scoring models.