[Paper] A Decision Support Framework for Blockchain Pattern Selection Based on Soft Goals

Source: arXiv - 2512.13239v1

Overview

The paper introduces BC‑TEAEM, a decision‑support framework that helps architects pick the right blockchain design patterns based on business‑level “soft goals” (e.g., scalability, privacy, regulatory compliance). By marrying a domain‑specific ontology of blockchain patterns with a multi‑criteria decision‑making (MCDM) process, the authors aim to make pattern selection systematic, transparent, and repeatable—something that’s often ad‑hoc in real‑world projects.

Key Contributions

• BC‑TEAEM framework that links high‑level business soft goals to concrete blockchain patterns through a formal ontology.
• Hybrid expert workflow that explicitly separates the domain expert (business side) from the technical expert (blockchain side) while ensuring continuous alignment and traceability.
• MCDM integration (using weighted scoring) to rank pattern alternatives according to the captured soft‑goal preferences.
• Prototype decision‑support tool that implements the framework and visualizes the selection process.
• Empirical validation via a case study on a pharmaceutical supply‑chain traceability system, demonstrating the framework’s practical applicability.

Methodology

1. Ontology Construction – The authors built two ontologies:

   • Blockchain Pattern Ontology: catalogues common patterns (e.g., token‑based access control, off‑chain storage, side‑chains).
   • Soft‑Goal Ontology: captures non‑functional business objectives such as “high throughput,” “low latency,” “regulatory auditability,” and “energy efficiency.”

2. Goal Elicitation & Weighting – A domain expert lists relevant soft goals for the project. The technical expert then assigns relative weights (e.g., using pairwise comparisons or a simple Likert scale) to express how critical each goal is.

3. Pattern Mapping – Each blockchain pattern is annotated with its impact on the soft goals (positive, neutral, or negative). This mapping is stored in the ontology, enabling automated reasoning.

4. Multi‑Criteria Decision Making – The framework computes a composite score for every pattern by aggregating the weighted soft‑goal impacts. Patterns are ranked, and the top‑scoring ones are presented to the architects.

5. Iterative Refinement – If the initial ranking is unsatisfactory, stakeholders can adjust goal weights or add/remove goals and re‑run the analysis, fostering a collaborative “design‑through‑discussion” loop.

Results & Findings

• In the pharmaceutical supply‑chain case, the framework identified a combination of private permissioned ledger, off‑chain data anchoring, and zero‑knowledge proof‑based privacy as the optimal pattern set, satisfying the company’s stringent regulatory auditability and data‑privacy requirements while keeping transaction costs low.
• The decision‑support tool reduced the pattern‑selection time from several weeks of ad‑hoc meetings to under a day of guided analysis.
• Stakeholders reported higher confidence in the chosen architecture because the reasoning process was explicitly documented and traceable back to business goals.

Practical Implications

• Accelerated Architecture Design – Development teams can quickly converge on a blockchain solution that aligns with business priorities, cutting down costly trial‑and‑error cycles.
• Risk Mitigation – By surfacing trade‑offs (e.g., privacy vs. scalability) early, organizations can avoid later re‑architecting or compliance penalties.
• Standardized Documentation – The ontology‑driven approach produces a reusable artifact that can be audited, shared across projects, or integrated into governance tools.
• Tool Integration – The prototype can be embedded into existing enterprise modeling platforms (e.g., ArchiMate, BPMN tools) or CI/CD pipelines to enforce pattern compliance automatically.
• Cross‑Domain Reusability – Although validated in pharma, the framework is domain‑agnostic; any industry that needs to balance soft goals (finance, IoT, logistics) can adopt it with minimal customization.

Limitations & Future Work

• Ontology Coverage – The current pattern catalog is limited to well‑known patterns; emerging constructs (e.g., DAG‑based ledgers, hybrid consensus mechanisms) are not yet modeled.
• Subjectivity in Weighting – Goal weights rely on expert judgment, which can introduce bias; the authors suggest exploring objective metrics or crowd‑sourced weighting in future studies.
• Scalability of the Tool – The prototype was tested on a single case study; performance and usability in large‑scale, multi‑project environments remain to be evaluated.
• Dynamic Goal Evolution – Real‑world projects often see soft goals evolve over time; extending the framework to support continuous re‑evaluation and versioning of goals is an open research direction.

Bottom line: BC‑TEAEM offers a pragmatic bridge between business intent and blockchain engineering, turning a traditionally fuzzy selection process into a transparent, repeatable, and collaborative workflow—exactly the kind of decision support that modern development teams need when navigating the rapidly expanding blockchain landscape.

Authors

• Eddy Kiomba Kambilo
• Nicolas Herbaut
• Irina Rychkova
• Carine Souveyet

Paper Information

• arXiv ID: 2512.13239v1
• Categories: cs.SE
• Published: December 15, 2025
• PDF: Download PDF