In this paper, we present SATGEN, the first implicit model that generates Boolean Satisfiability formulas which resemble instances that arise in real-world industrial settings. Our approach uses unsupervised machine learning techniques to create new formulas by mimicking the structural properties of a given input formula Φ. We proceed in two phases: first, we construct the Literal Incidence Graph (LIG) of Φ. This is used by a Generative Adversarial Network to generate new LIGs that exhibit graph-theoretic properties similar to those of the LIG of Φ. In the second phase, we extract a formula Φ′ whose LIG would correspond to the generated graph. Generating such a formula is equivalent to finding a minimal clique edge cover of the given graph, which we tackle efficiently using a greedy hill-climbing algorithm. We verify experimentally that our approach generates formulas that closely resemble a given real-world SAT instance, as measured by a range of different metrics.
[1]
Bart Selman,et al.
Ten Challenges in Propositional Reasoning and Search
,
1997,
IJCAI.
[2]
Thomas Stützle,et al.
SATLIB: An Online Resource for Research on SAT
,
2000
.
[3]
Armin Biere,et al.
Effective Preprocessing in SAT Through Variable and Clause Elimination
,
2005,
SAT.
[4]
Maria Luisa Bonet,et al.
On the Structure of Industrial SAT Instances
,
2009,
CP.
[5]
Jesús Giráldez-Cru,et al.
A Modularity-Based Random SAT Instances Generator
,
2015,
IJCAI.
[6]
Jesús Giráldez-Cru,et al.
Locality in Random SAT Instances
,
2017,
IJCAI.
[7]
Stephan Günnemann,et al.
NetGAN: Generating Graphs via Random Walks
,
2018,
ICML.