Hypergraph Neural Networks Accelerate MUS Enumeration
This addresses the challenge of exponential search space growth in MUS enumeration, particularly for domains beyond Boolean satisfiability, though it is incremental as it builds on existing machine learning approaches.
The paper tackled the problem of accelerating Minimal Unsatisfiable Subsets (MUSes) enumeration in constraint satisfaction problems by proposing a domain-agnostic method using Hypergraph Neural Networks (HGNNs), which reduces the number of satisfiability checks needed and enumerates more MUSes within the same budget compared to conventional methods.
Enumerating Minimal Unsatisfiable Subsets (MUSes) is a fundamental task in constraint satisfaction problems (CSPs). Its major challenge is the exponential growth of the search space, which becomes particularly severe when satisfiability checks are expensive. Recent machine learning approaches reduce this cost for Boolean satisfiability problems but rely on explicit variable-constraint relationships, limiting their application domains. This paper proposes a domain-agnostic method to accelerate MUS enumeration using Hypergraph Neural Networks (HGNNs). The proposed method incrementally builds a hypergraph with constraints as vertices and MUSes enumerated until the current step as hyperedges, and employs an HGNN-based agent trained via reinforcement learning to minimize the number of satisfiability checks required to obtain an MUS. Experimental results demonstrate the effectiveness of our approach in accelerating MUS enumeration, showing that our method can enumerate more MUSes within the same satisfiability check budget compared to conventional methods.