Machine Generalization and Human Categorization: An Information-Theoretic View

TL;DR

This paper proposes an information-theoretic approach to machine generalization and categorization, aligning computational models with human psychological data to improve system explanations and generalizations.

Contribution

It introduces an information-based measure of category value that better predicts human categorization behavior than traditional methods.

Findings

Information measure predicts categorization experiments more accurately.

Aligns machine generalization with human cognitive patterns.

Supports natural and understandable AI explanations.

Abstract

In designing an intelligent system that must be able to explain its reasoning to a human user, or to provide generalizations that the human user finds reasonable, it may be useful to take into consideration psychological data on what types of concepts and categories people naturally use. The psychological literature on concept learning and categorization provides strong evidence that certain categories are more easily learned, recalled, and recognized than others. We show here how a measure of the informational value of a category predicts the results of several important categorization experiments better than standard alternative explanations. This suggests that information-based approaches to machine generalization may prove particularly useful and natural for human users of the systems.