Emerging Dimension Weights in a Conceptual Spaces Model of Concept Combination

TL;DR

This paper introduces a new framework for combining concepts in a conceptual space model, demonstrating how shared weights emerge in language evolution simulations influenced by environmental factors and social learning rates.

Contribution

The paper presents a novel semantic framework for conjunctive concept combination and shows how shared weights evolve in multi-agent simulations based on environmental and social parameters.

Findings

Shared weights emerge in language evolution simulations.

Variance of weights depends on environmental distribution and adoption rate.

Convergence to steady state takes longer with lower adoption rates.

Abstract

We investigate the generation of new concepts from combinations of properties as an artificial language develops. To do so, we have developed a new framework for conjunctive concept combination. This framework gives a semantic grounding to the weighted sum approach to concept combination seen in the literature. We implement the framework in a multi-agent simulation of language evolution and show that shared combination weights emerge. The expected value and the variance of these weights across agents may be predicted from the distribution of elements in the conceptual space, as determined by the underlying environment, together with the rate at which agents adopt others' concepts. When this rate is smaller, the agents are able to converge to weights with lower variance. However, the time taken to converge to a steady state distribution of weights is longer.