DeepProofLog: Efficient Proving in Deep Stochastic Logic Programs

2025-11-11

AI Summary

Neural-symbolic (NeSy) systems offer interpretability and strong generalization but suffer from severe scalability bottlenecks in subsymbolic logical reasoning. To address this, we propose DeepProofLog (DPrL), a novel NeSy framework grounded in stochastic logic programs. DPrL establishes, for the first time, a formal theoretical mapping between deep logical deduction and Markov decision processes, framing proof search as a joint optimization problem of dynamic programming and reinforcement learning. It parameterizes inference steps via neural networks, integrating neural guidance with stochastic logic programs to enable efficient, scalable, end-to-end learning. Evaluated on standard NeSy benchmarks and knowledge graph reasoning tasks, DPrL substantially outperforms state-of-the-art methods, achieving significant improvements in both reasoning scale and computational efficiency.

Technology Category

• Machine Learning: Neuro-Symbolic Learning
• Knowledge Representation and Reasoning: Logic Programming
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

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.

Problem

• Addresses scalability limitations in neurosymbolic AI systems
• Enables efficient neural guidance over logical proving systems
• Improves inference scalability for complex proof spaces and knowledge bases

Innovation

• Neural networks parameterize all derivation steps
• Maps resolution process to Markov Decision Processes
• Applies dynamic programming and reinforcement learning