CLARK: Closed-loop Learning for Adaptive Reasoning over Knowledge Graphs

2026-07-22

AI Summary

This work addresses the limitations of traditional machine learning in handling distribution shifts and integrating prior knowledge with uncertainty-aware reasoning. To overcome these challenges, the authors propose CLARK, a novel framework that introduces, for the first time, a closed-loop learning mechanism for adaptive reasoning over knowledge graphs. CLARK iteratively refines logical rules and their probabilistic weights by combining symbolic rule mining with Markov Logic Networks (LP$^{\text{MLN}}$) on knowledge graphs constructed via CACTUS. The approach achieves substantial improvements in classification performance and out-of-distribution generalization while preserving high interpretability. Empirical evaluation on two medical datasets demonstrates CLARK’s effectiveness, yielding logical rules that are both accurate and semantically transparent.

Abstract

Machine Learning models are widely used for automating classification tasks by extracting statistical patterns from data. However, their performance deteriorates if the data distribution changes, making them ill-suited to handle uncertain and evolving information. Moreover, they provide limited support for integrating prior knowledge. To address these limitations, we present CLARK (Closed-loop Learning for Adaptive Reasoning over Knowledge Graphs), a framework that integrates knowledge graphs, symbolic rule mining, and probabilistic reasoning under the Logic Programs with Markov Logic Networks (LP$^{\text{MLN}}$) formalism. Starting from CACTUS-derived KGs, CLARK translates graph structure into an LP$^{\text{MLN}}$ program and iteratively enriches it with candidate rules proposed by symbolic learners. These rules are calibrated through probabilistic weight learning, enabling reasoning under uncertainty and refinement of the underlying graph structure. We evaluate CLARK on two medical datasets, analysing both rule quality and downstream classification performance. Results demonstrate that CLARK leads to improved classification performance and more generalisable inference. Overall, CLARK provides a principled approach to constructing adaptive, interpretable, knowledge-driven models for classification.

Problem

• distribution shift
• uncertain information
• evolving data
• prior knowledge integration
• knowledge-driven reasoning

Innovation

• Knowledge Graphs
• Markov Logic Networks
• Symbolic Rule Mining
• Closed-loop Learning
• Probabilistic Reasoning