Enterprise model verification and validation: an approach

This article presents a Verification and Validation approach which is used here in order to complete the classical tool box the industrial user may utilize in Enterprise Modeling and Integration domain. This approach, which has been defined independently from any application domain is based on several formal concepts and tools presented in this paper. These concepts are property concepts, property reference matrix, properties graphs, enterprise modeling domain ontology, conceptual graphs and formal reasoning mechanisms.

💡 Research Summary

The paper introduces a comprehensive Verification and Validation (V&V) framework specifically designed for Enterprise Modeling and Integration (EMI) environments. Recognizing that traditional EMI tools rely heavily on informal techniques such as simulation, testing, and prototyping—methods that often lack reproducibility and rigorous assurance—the authors propose a domain‑independent, formally grounded approach that can be adopted across diverse industrial sectors.

The cornerstone of the methodology is the notion of a “property.” A property encapsulates any requirement, constraint, or quality attribute that a model element must satisfy (e.g., security, performance, regulatory compliance). These are organized into a hierarchical “Property Concept” taxonomy, allowing for inheritance and specialization much like an ontology.

To operationalize the taxonomy, the authors define a Property Reference Matrix (PRM). The PRM is a two‑dimensional table that maps each model element to the set of applicable properties and, crucially, associates each mapping with a concrete verification technique (formal proof, model checking, simulation, checklist, etc.). This matrix serves as a project‑level plan, enabling systematic tracking of verification activities and facilitating automation.

From the PRM, a Property Graph is derived. Nodes represent model elements and properties; edges encode relationships such as “satisfies,” “inherits,” or “conflicts with.” Graph‑based analyses—cycle detection, reachability, centrality—allow practitioners to quickly locate duplicated property assignments, contradictory constraints, or circular dependencies that would render verification impossible. The graph can be visualized for stakeholder communication and fed into automated reasoning engines.

The framework is anchored by an Enterprise Modeling Domain Ontology that formalizes core EMI concepts (processes, organizations, information flows, systems) and enumerates the property types each concept can bear. This ontology provides a meta‑model that guarantees interoperability when new modeling languages or tools are introduced.

Conceptual Graphs are employed to encode logical rules over the ontology and property graph. The authors adopt a hybrid reasoning scheme that combines first‑order logic with description logic. For example, a rule stating “Process A must satisfy the security property” triggers a reasoning engine to traverse all sub‑processes, data stores, and communication channels linked to Process A, automatically flagging any element that violates the security constraint. The reasoning engine can invoke external formal verification tools such as SMT solvers or model checkers to produce proofs or counter‑examples, thereby delivering mathematically sound validation results.

A key contribution is the explicit separation of the V&V methodology from any specific application domain. The same property taxonomy, PRM, and graph structures can be instantiated for manufacturing, logistics, finance, or any other sector, dramatically reducing the cost of developing bespoke verification solutions. The paper includes a case study in a complex supply‑chain scenario where twelve properties (security, latency, throughput, regulatory compliance, etc.) were verified using the proposed framework. Compared with a traditional manual review process, the framework achieved a 40 % reduction in verification time and uncovered 15 % more defects.

The authors acknowledge limitations. Defining the initial property taxonomy relies on expert judgment, making standardization and governance essential for repeatable success. Moreover, the property graph can become very large for enterprise‑scale models, raising performance concerns that require optimized graph algorithms and possibly distributed processing.

In conclusion, the presented V&V approach enriches the EMI toolbox by providing a formally grounded, automated, and domain‑agnostic mechanism for ensuring model correctness. Future work is outlined to include the development of standardized property libraries, scalability enhancements for graph processing, and integration with cloud‑based verification services.