Artifact systems with data dependencies and arithmetic

Abstract

We study the static verification problem for data-centric business processes, specified in a variant of IBM's “business artifact” model. Artifacts are records of variables that correspond to business-relevant objects and are updated by a set of services equipped with pre- and postconditions, that implement business process tasks. The verification problem consists in statically checking whether all runs of an artifact system satisfy desirable properties expressed in a first-order extension of linear-time temporal logic. Previous work identified the class of guarded artifact systems and properties, for which verification is decidable. However, the results suffer an important limitation: they fail in the presence of even very simple data dependencies or arithmetic, both crucial to real-life business processes. In this article, we extend the artifact model and verification results to alleviate this limitation. We identify a practically significant class of business artifacts with data dependencies and arithmetic, for which verification is decidable. The technical machinery needed to establish the results is fundamentally different from previous work. While the worst-case complexity of verification is nonelementary, we identify various realistic restrictions yielding more palatable upper bounds.

Recency-Bounded Verification of Dynamic Database-Driven Systems

“…We also aim at applying under-approximation techniques in the case where the initial database is not known, and model checking is studied for every possible initial database, in the style of [14,12,8]. A transition may be explicitly triggered by an action applied to the artifact instance, or implicitly because another artifact instance is entering into a specific state (this dependency is rendered using dahshed arrows).…”

“…Example 4.3. The presence of database constraints in the dynamic system under study is a key feature, which has been extensively studied in the literature [15,14,12,8,5]. In our setting, arbitrary FO constraints can be seamlessly added, adopting the semantics, as in [5], that the application of an action is blocked whenever the resulting database instance violates one of the constraints.…”

“…Two main trends can be identified along this line. The first trend was initiated in the late 1990s with the introduction of relational transducers [2], and continued with new results over progressively richer variants of the initial model, such as systems equipped with arithmetic [14,12], systems decomposed into interacting web services [15], and systems operating over XML databases [8]. The modelling formalisms introduced in this direction operate over a read-only, input database that is fixed during the system evolution, and use quantifier-free FO formulae to query such a database.…”

“…The obtained answers can be stored into a read-write state database, whose size is fixed a-priori. Verification problems include control-state reachability [8], or model checking [14,12] against formulae expressed in FO variants of temporal logics with a limited form of FO quantification across state. Furthermore, verification is input-parametric, that is, studied independently from the configuration of data in the initial input database.…”

Verification of Time-Aware Business Processes Using Constrained Horn Clauses

“…However, the above mentioned formalisms and tools may not be adequate if we want to complement time-based reasoning with general purpose logical reasoning, which is often needed if we take into account more complex aspects of knowledge manipulation activities relative to business processes. For instance, some verification approaches make use of ontology-based reasoning about the business domain where processes are executed [27,30], while others combine reasoning on the finite-state process behavior with reasoning on the manipulation of data objects of an infinite type, such as databases or integers [8,2,26].…”