Large language models can flag contradictions and errors in business documentation and extract process structures, but the reported evidence is thin, partly synthetic, and lacks ground-truth metrics for the hardest task.
Formal Modeling and SMT-Based Parameterized Verification of Data-Aware BPMN (Extended Version)
1 Pith paper cite this work, alongside 3 external citations. Polarity classification is still indexing.
1
Pith paper citing it
3
external citations · Pith
abstract
We propose DAB -- a data-aware extension of BPMN where the process operates over case and persistent data (partitioned into a read-only database called catalog and a read-write database called repository). The model trades off between expressiveness and the possibility of supporting parameterized verification of safety properties on top of it. Specifically, taking inspiration from the literature on verification of artifact systems, we study verification problems where safety properties are checked irrespectively of the content of the read-only catalog, and accepting the potential presence of unboundedly many tuples in the catalog and repository. We tackle such problems using an array-based backward reachability procedure fully implemented in MCMT -- a state-of-the-art array-based SMT model checker. Notably, we prove that the procedure is sound and complete for checking safety of DABs, and single out additional conditions that guarantee its termination and, in turn, show decidability of checking safety.
fields
cs.SE 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
LLM-based Vulnerability Discovery in Business Process Documentation
Large language models can flag contradictions and errors in business documentation and extract process structures, but the reported evidence is thin, partly synthetic, and lacks ground-truth metrics for the hardest task.