TY - GEN
T1 - Online enumeration of all minimal inductive validity cores
AU - Bendík, Jaroslav
AU - Ghassabani, Elaheh
AU - Whalen, Michael
AU - Černá, Ivana
PY - 2018/1/1
Y1 - 2018/1/1
N2 - Symbolic model checkers can construct proofs of safety properties over complex models, but when a proof succeeds, the results do not generally provide much insight to the user. Minimal Inductive Validity Cores (MIVCs) trace a property to a minimal set of model elements necessary for constructing a proof, and can help to explain why a property is true of a model. In addition, the traceability information provided by MIVCs can be used to perform a variety of engineering analysis such as coverage analysis, robustness analysis, and vacuity detection. The more MIVCs are identified, the more precisely such analyses can be performed. Nevertheless, a full enumeration of all MIVCs is in general intractable due to the large number of possible model element sets. The bottleneck of existing algorithms is that they are not guaranteed to emit minimal IVCs until the end of the computation, so returned results are not known to be minimal until all solutions are produced. In this paper, we propose an algorithm that identifies MIVCs in an online manner (i.e., one by one) and can be terminated at any time. We benchmark our new algorithm against existing algorithms on a variety of examples, and demonstrate that our algorithm not only is better in intractable cases but also completes the enumeration of MIVCs faster than competing algorithms in many tractable cases.
AB - Symbolic model checkers can construct proofs of safety properties over complex models, but when a proof succeeds, the results do not generally provide much insight to the user. Minimal Inductive Validity Cores (MIVCs) trace a property to a minimal set of model elements necessary for constructing a proof, and can help to explain why a property is true of a model. In addition, the traceability information provided by MIVCs can be used to perform a variety of engineering analysis such as coverage analysis, robustness analysis, and vacuity detection. The more MIVCs are identified, the more precisely such analyses can be performed. Nevertheless, a full enumeration of all MIVCs is in general intractable due to the large number of possible model element sets. The bottleneck of existing algorithms is that they are not guaranteed to emit minimal IVCs until the end of the computation, so returned results are not known to be minimal until all solutions are produced. In this paper, we propose an algorithm that identifies MIVCs in an online manner (i.e., one by one) and can be terminated at any time. We benchmark our new algorithm against existing algorithms on a variety of examples, and demonstrate that our algorithm not only is better in intractable cases but also completes the enumeration of MIVCs faster than competing algorithms in many tractable cases.
KW - Inductive proofs
KW - Inductive validity cores
KW - Proof cores
KW - SMT-based model checking
KW - Traceability
UR - http://www.scopus.com/inward/record.url?scp=85049047738&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85049047738&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-92970-5_12
DO - 10.1007/978-3-319-92970-5_12
M3 - Conference contribution
AN - SCOPUS:85049047738
SN - 9783319929699
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 189
EP - 204
BT - Software Engineering and Formal Methods - 16th International Conference, SEFM 2018, Held as Part of STAF 2018, Proceedings
A2 - Johnsen, Einar Broch
A2 - Schaefer, Ina
PB - Springer- Verlag
T2 - 16th International Conference on Software Engineering and Formal Methods, SEFM 2018 Held as Part of STAF 2018
Y2 - 27 June 2018 through 29 June 2018
ER -