Distributional Measures of Semantic Distance: A Survey

TL;DR

This survey reviews distributional and WordNet-based measures of semantic distance, analyzing their strengths, limitations, and potential for aligning with human semantic judgments, including recent hybrid approaches.

Contribution

It provides a comprehensive comparison of distributional and knowledge-based semantic distance measures and discusses ways to improve their alignment with human perception.

Findings

Distributional measures are useful in resource-poor languages.

WordNet-based measures generally align better with human judgment.

Hybrid approaches show promise for improved semantic distance estimation.

Abstract

The ability to mimic human notions of semantic distance has widespread applications. Some measures rely only on raw text (distributional measures) and some rely on knowledge sources such as WordNet. Although extensive studies have been performed to compare WordNet-based measures with human judgment, the use of distributional measures as proxies to estimate semantic distance has received little attention. Even though they have traditionally performed poorly when compared to WordNet-based measures, they lay claim to certain uniquely attractive features, such as their applicability in resource-poor languages and their ability to mimic both semantic similarity and semantic relatedness. Therefore, this paper presents a detailed study of distributional measures. Particular attention is paid to flesh out the strengths and limitations of both WordNet-based and distributional measures, and how distributional measures of distance can be brought more in line with human notions of semantic distance. We conclude with a brief discussion of recent work on hybrid measures.