ExplainabilityTransparencyAccountabilitycs.LO
Bridging Logic Programming and Deep Learning for Explainability through ILASP
Talissa Dreossi
February 13, 2025
Electronic Proceedings in Theoretical Computer Science
Key Takeaways
-
Hybrid AI systems combining neural networks with logic programming can generate both accurate predictions and human-readable explanations, addressing a key limitation of traditional deep learning where decisions remain opaque "black boxes."
-
The research demonstrates practical applications across high-stakes domains including legal decision-making (interpreting Italian court decisions), weather forecasting, and medical/biological classification (spermatozoa morphology), suggesting this approach is particularly valuable where explainability is critical for professional trust and accountability.
-
Logic-based explanations derived through Inductive Logic Programming (ILP) can validate neural network conclusions, providing a verification mechanism that helps ensure AI systems aren't just accurate but also reasoning in defensible ways—important for responsible deployment in domains like law and healthcare.
-
This approach addresses a fundamental tension in responsible AI: the need for both performance (where deep learning excels) and transparency (where traditional logic-based systems excel), rather than forcing a choice between them.
Abstract
My research explores integrating deep learning and logic programming to set the basis for a new generation of AI systems. By combining neural networks with Inductive Logic Programming (ILP), the goal is to construct systems that make accurate predictions and generate comprehensible rules to validate these predictions. Deep learning models process and analyze complex data, while ILP techniques derive logical rules to prove the network's conclusions. Explainable AI methods, like eXplainable Answer Set Programming (XASP), elucidate the reasoning behind these rules and decisions. The focus is on applying ILP frameworks, specifically ILASP and FastLAS, to enhance explainability in various domains. My test cases span weather prediction, the legal field, and image recognition. In weather forecasting, the system will predict events and provides explanations using FastLAS, with plans to integrate recurrent neural networks in the future. In the legal domain, the research focuses on interpreting vague decisions and assisting legal professionals by encoding Italian legal articles and learning reasoning patterns from Court of Cassation decisions using ILASP. For biological laboratories, we will collaborate with a research group to automate spermatozoa morphology classification for Bull Breeding Soundness Evaluation using YOLO networks and ILP to explain classification outcomes. This hybrid approach aims to bridge the gap between the high performance of deep learning models and the transparency of symbolic reasoning, advancing AI by providing interpretable and trustworthy applications.