DeepProofLog: Efficient Proving in Deep Stochastic Logic Programs

TL;DR

DeepProofLog (DPrL) is a neurosymbolic system that combines neural networks with stochastic logic programming to enable scalable, efficient reasoning and learning in complex AI tasks, outperforming existing methods.

Contribution

DPrL introduces neural-guided stochastic logic programming with a formal MDP mapping, significantly improving scalability and inference efficiency in neurosymbolic AI.

Findings

DPrL outperforms state-of-the-art NeSy systems on benchmarks.

The formal MDP mapping enables efficient dynamic programming techniques.

DPrL handles larger, more complex knowledge bases effectively.

Abstract

Neurosymbolic (NeSy) AI aims to combine the strengths of neural architectures and symbolic reasoning to improve the accuracy, interpretability, and generalization capability of AI models. While logic inference on top of subsymbolic modules has been shown to effectively guarantee these properties, this often comes at the cost of reduced scalability, which can severely limit the usability of NeSy models. This paper introduces DeepProofLog (DPrL), a novel NeSy system based on stochastic logic programs, which addresses the scalability limitations of previous methods. DPrL parameterizes all derivation steps with neural networks, allowing efficient neural guidance over the proving system. Additionally, we establish a formal mapping between the resolution process of our deep stochastic logic programs and Markov Decision Processes, enabling the application of dynamic programming and reinforcement learning techniques for efficient inference and learning. This theoretical connection improves scalability for complex proof spaces and large knowledge bases. Our experiments on standard NeSy benchmarks and knowledge graph reasoning tasks demonstrate that DPrL outperforms existing state-of-the-art NeSy systems, advancing scalability to larger and more complex settings than previously possible.