Bridging the Gap: Representation Spaces in Neuro-Symbolic AI

TL;DR

This paper analyzes the differences in data representation between neural networks and symbolic learning in neuro-symbolic AI, proposing a classification framework to understand their integration and collaboration strategies.

Contribution

It introduces a four-level classification framework for representation spaces, modalities, logic methods, and collaboration strategies in neuro-symbolic AI.

Findings

Analyzed 191 studies to categorize representation spaces and modalities.

Detailed analysis of 46 research works based on their representation space.

Provides insights into collaboration strategies between neural and symbolic methods.

Abstract

Neuro-symbolic AI is an effective method for improving the overall performance of AI models by combining the advantages of neural networks and symbolic learning. However, there are differences between the two in terms of how they process data, primarily because they often use different data representation methods, which is often an important factor limiting the overall performance of the two. From this perspective, we analyzed 191 studies from 2013 by constructing a four-level classification framework. The first level defines five types of representation spaces, and the second level focuses on five types of information modalities that the representation space can represent. Then, the third level describes four symbolic logic methods. Finally, the fourth-level categories propose three collaboration strategies between neural networks and symbolic learning. Furthermore, we conducted a detailed analysis of 46 research based on their representation space.