Machine Learning, Clustering, and Polymorphy

TL;DR

This paper introduces WITT, a machine induction program that models human categorization by capturing properties like prototypicality, contrasts, and polymorphy, offering an alternative to traditional AI clustering methods.

Contribution

The paper presents WITT, a novel approach to modeling human categorization that emphasizes polymorphy and contrasts, differing from conventional feature-based clustering techniques.

Findings

WITT aligns more closely with human categorization patterns.

WITT can replicate results of traditional clustering schemes.

Applications in expert systems and information retrieval are discussed.

Abstract

This paper describes a machine induction program (WITT) that attempts to model human categorization. Properties of categories to which human subjects are sensitive includes best or prototypical members, relative contrasts between putative categories, and polymorphy (neither necessary or sufficient features). This approach represents an alternative to usual Artificial Intelligence approaches to generalization and conceptual clustering which tend to focus on necessary and sufficient feature rules, equivalence classes, and simple search and match schemes. WITT is shown to be more consistent with human categorization while potentially including results produced by more traditional clustering schemes. Applications of this approach in the domains of expert systems and information retrieval are also discussed.