The Pet-Fish problem on the World-Wide Web

TL;DR

This paper investigates the Pet-Fish problem and the Guppy effect on the Web by introducing weights and 'meaning bound' measures to analyze concept combinations and their typical exemplars.

Contribution

It presents a novel method using absolute and relative weights, along with 'meaning bound' metrics, to analyze concept interactions on the Web, highlighting the Pet-Fish problem.

Findings

Identification of Pet-Fish problem situations on the Web

Introduction of 'meaning bound' as a measure of concept association

Demonstration of stronger conjunctions than individual concepts

Abstract

We identify the presence of Pet-Fish problem situations and the corresponding Guppy effect of concept theory on the World-Wide Web. For this purpose, we introduce absolute weights for words expressing concepts and relative weights between words expressing concepts, and the notion of 'meaning bound' between two words expressing concepts, making explicit use of the conceptual structure of the World-Wide Web. The Pet-Fish problem occurs whenever there are exemplars - in the case of Pet and Fish these can be Guppy or Goldfish - for which the meaning bound with respect to the conjunction is stronger than the meaning bounds with respect to the individual concepts.