[Paper Review] Process Query Language: Design, Implementation, and Evaluation

Artem Polyvyanyy, Arthur H. M. ter Hofstede+3 more|arXiv (Cornell University)|Sep 20, 2019

Business Process Modeling and Analysis4 citations

TL;DR

This paper introduces Process Query Language (PQL), a domain-specific language for querying process model repositories based on behavioral relations between tasks, such as ordering, concurrency, or exclusion. The language uses empirically validated behavioral predicates with formally proven decidability and demonstrates near real-time performance on large-scale model collections, enabling effective auditing and automation through behavior-aware querying.

ABSTRACT

Organizations can benefit from the use of practices, techniques, and tools from the area of business process management. Through the focus on processes, they create process models that require management, including support for versioning, refactoring and querying. Querying thus far has primarily focused on structural properties of models rather than on exploiting behavioral properties capturing aspects of model execution. While the latter is more challenging, it is also more effective, especially when models are used for auditing or process automation. The focus of this paper is to overcome the challenges associated with behavioral querying of process models in order to unlock its benefits. The first challenge concerns determining decidability of the building blocks of the query language, which are the possible behavioral relations between process tasks. The second challenge concerns achieving acceptable performance of query evaluation. The evaluation of a query may require expensive checks in all process models, of which there may be thousands. In light of these challenges, this paper proposes a special-purpose programming language, namely Process Query Language (PQL) for behavioral querying of process model collections. The language relies on a set of behavioral predicates between process tasks, whose usefulness has been empirically evaluated with a pool of process model stakeholders. This study resulted in a selection of the predicates to be implemented in PQL, whose decidability has also been formally proven. The computational performance of the language has been extensively evaluated through a set of experiments against two large process model collections.

Motivation & Objective

• To address the gap in existing process query languages that focus on structural rather than behavioral aspects of process models.
• To enable efficient and semantically meaningful querying of process model repositories based on behavioral relations between tasks.
• To empirically validate the relevance of behavioral predicates for process model querying through stakeholder studies.
• To ensure decidability of behavioral predicates through formal proofs, enhancing reliability in query evaluation.
• To evaluate the performance of PQL on real and synthetic process model collections to demonstrate feasibility and scalability.

Proposed method

• Designing PQL with an SQL-like syntax to improve usability and support complex behavioral queries.
• Defining a set of core behavioral predicates (e.g., can-execute, never-execute, concurrent) based on empirical evaluation with process model stakeholders.
• Proving the decidability of each behavioral predicate formally to ensure correctness and consistency in query evaluation.
• Implementing a query engine that evaluates PQL queries over large process model repositories using optimized algorithms.
• Supporting label similarity matching with a configurable threshold to handle naming variations in process tasks.
• Evaluating performance using real-world and synthetic process model collections to measure query execution time and scalability.

Figure 3. Abstract syntax tree of the example PQL query Q3 in Section 2 . The numbers in parenthesis are used in Section 4.5 to explain the meaning of the query.

Experimental results

Research questions

• RQ1 Which behavioral relations between process tasks are most relevant and useful for querying process model repositories?
• RQ2 Can behavioral predicates be formally proven decidable to ensure reliable query evaluation?
• RQ3 What is the performance of a query language based on behavioral predicates when applied to large-scale process model collections?
• RQ4 How does PQL support intuitive and effective querying of process models through label similarity and behavior-aware semantics?
• RQ5 To what extent can PQL support complex search intents, such as pattern-based queries with wildcards?

Key findings

• The empirical study identified a set of behaviorally meaningful predicates that are most relevant for process model querying, forming the foundation of PQL.
• All core PQL predicates were formally proven decidable, ensuring correctness and enabling reliable query evaluation.
• The PQL query engine achieved near real-time response times on large-scale process model collections, demonstrating practical feasibility.
• Label similarity matching with a configurable threshold effectively handles naming variations across models, improving retrieval accuracy.
• Performance evaluation confirmed that PQL can scale to thousands of models with acceptable response times, supporting industrial use cases.
• The language supports complex search patterns, including wildcard-based execution templates, enabling expressive behavioral queries.

Figure 6. A system obtained via unification of the set of labels $\{\texttt{a},\texttt{a}^{\prime}\}$ in the workflow system in Figure 4 .