DIAPROD Summative Validation: Greenshire Council Public Sector Case Study Instance

Summative validation artefacts for the DIAPROD (Dynamic DataProtection by Design) framework, evaluating the complete ConsentMLmetamodel — including the data minimisation and storage limitationextensions — against a public sector scenario not seen duringframework development (FEDS Quadrant 4: summative, ex post). This deposit contains: (i) a graphical model of the GreenshireCouncil public sector case study constructed in the DIAPROT(Dynamic Data Protection Tool), containing base ConsentMLconstructs (five actors, three consents, four legal bases, threeprocessing purposes, nine personal data elements, one datacategory, three policies) connected to extension constructs (threepurpose–data linkages, two data scopes, three retention policies,five retention states, one erasure trigger, one erasure candidate,and two retention audit records) through the extensionrelationships; and (ii) a machine-readable JSON instance of thecomplete case study model (46 nodes, 136 edges, 18 node types,35 edge types) exported from the DIAPROT tool in ConsentML-JSONformat for programmatic validation and reproducibility. The case study evaluates six competency questions testing featuresnot exercised by the ZipFin FinTech (formative) or BritPayfinancial services (summative) case studies: (CQ-P1) non-consentlegal basis override — email_address retained under council tax(Art. 6(1)(e), public task, 7-year) when newsletter consent iswithdrawn, because the public task basis cannot be withdrawn;(CQ-P2) purpose completion with statutory retention — Alice Patelmoves out of the borough, but council tax data is retained for 7years under the Limitation Act 1980; (CQ-P3) permanent archival —planning application records archived permanently under the Townand Country Planning Act 1990 with Art. 89(1) safeguards, usingthe five-state lifecycle's 'archived' terminal state; (CQ-P4)regulatory evolution — a court ruling changes the legal basis forcouncil tax processing from public task (Art. 6(1)(e)) tolegitimate interest (Art. 6(1)(f)), requiring only a metadataupdate with no metamodel modification; (CQ-P5) contextual dataclassification with special category data — disability_status(Art. 9) required only when γ = 'council tax reductionapplication'; and (CQ-P6) regulatory inspection support — allpredicates (C-MIN-1..8, C-SL-1..8) produce an exportablevalidation report for ICO audit. The artefacts support the doctoral research by demonstrating thatthe same 34-class metamodel, 82 typed relationships, and 58well-formedness predicates validated the public sector domainwithout metamodel modification — confirming the framework'sgeneralisability across three domains (FinTech, financialservices, public sector). The Greenshire scenario uniquelyexercises: four layered legal bases (Art. 6(1)(a)(c)(e)(f)),non-consent processing as the majority legal basis, permanentarchival (the 'archived' retention state), regulatory evolution(legal basis change without architectural modification), specialcategory contextual data (disability_status under Art. 9), threegoverning policies (GDPR, LGFA 1992, TCPA 1990), and all fiveretention lifecycle states (active, expired, held, erased,archived) within a single case study. The extension integrates with ConsentML, a metamodel-basedmodelling language for dynamic consent within the DIAPRODframework, described in: Daniels, S.O., Mouratidis, H.:ConsentML: A Modelling Language for Dynamic Consent Modelling. In:RCIS 2026, LNBIP, vol. 585, pp. 542–557. Springer (2026).https://doi.org/10.1007/978-3-032-26836-5_33 All materials are provided in open formats for reproducibility.