Query Stability in Monotonic Data-Aware Business Processes [Extended Version]

TL;DR

This paper studies the stability of conjunctive queries in business processes that continuously generate data, providing formal methods to determine when query answers remain consistent despite ongoing data changes.

Contribution

It introduces a formal framework combining process control flow with data read/write specifications and characterizes query stability under various process restrictions, including decidability and optimal encodings.

Findings

Decidability results for query stability under different process restrictions

Formal encodings into Datalog variants with optimal complexity

Identification of process facets affecting stability

Abstract

Organizations continuously accumulate data, often according to some business processes. If one poses a query over such data for decision support, it is important to know whether the query is stable, that is, whether the answers will stay the same or may change in the future because business processes may add further data. We investigate query stability for conjunctive queries. To this end, we define a formalism that combines an explicit representation of the control flow of a process with a specification of how data is read and inserted into the database. We consider different restrictions of the process model and the state of the system, such as negation in conditions, cyclic executions, read access to written data, presence of pending process instances, and the possibility to start fresh process instances. We identify for which facet combinations stability of conjunctive queries is decidable and provide encodings into variants of Datalog that are optimal with respect to the worst-case complexity of the problem.