Unsupervised Neural-Symbolic Integration

TL;DR

This paper presents a novel approach to integrating symbolic knowledge into unsupervised neural networks, enhancing their reasoning and interpretability capabilities across various knowledge representations.

Contribution

It introduces methods for unsupervised neural-symbolic integration, extending previous supervised-focused work to unsupervised models with diverse symbolic knowledge forms.

Findings

Successful integration of propositional logic for DNA promoter prediction

Application of first-order logic to understand family relationships

Enhanced interpretability of neural networks through symbolic knowledge

Abstract

Symbolic has been long considered as a language of human intelligence while neural networks have advantages of robust computation and dealing with noisy data. The integration of neural-symbolic can offer better learning and reasoning while providing a means for interpretability through the representation of symbolic knowledge. Although previous works focus intensively on supervised feedforward neural networks, little has been done for the unsupervised counterparts. In this paper we show how to integrate symbolic knowledge into unsupervised neural networks. We exemplify our approach with knowledge in different forms, including propositional logic for DNA promoter prediction and first-order logic for understanding family relationship.