A Comparison of WordNet and Roget's Taxonomy for Measuring Semantic Similarity

TL;DR

This paper evaluates the effectiveness of Roget's Thesaurus as a taxonomy for semantic similarity measurement, comparing it with WordNet and traditional edge counting methods, showing promising results close to human judgments.

Contribution

It introduces the use of Roget's Thesaurus for semantic similarity measurement and compares its performance with existing methods like WordNet.

Findings

Edge counting with Roget's achieves a correlation of r=0.88 with human judgments.

Roget's Thesaurus performs comparably to WordNet in semantic similarity tasks.

Traditional edge counting is surprisingly effective for measuring semantic similarity.

Abstract

This paper presents the results of using Roget's International Thesaurus as the taxonomy in a semantic similarity measurement task. Four similarity metrics were taken from the literature and applied to Roget's The experimental evaluation suggests that the traditional edge counting approach does surprisingly well (a correlation of r=0.88 with a benchmark set of human similarity judgements, with an upper bound of r=0.90 for human subjects performing the same task.)