PhD Dissertation · Chapter 5 · Serialised Standard Corpus

SDA Standards Explorerv2.2 · W6.1

About this instrument

SDA Standards Explorer

A governed, self-contained instrument over the serialised NDIS SDA Design Standard 2019 corpus, produced for Chapter 5 of the dissertation. Two channels are represented: the text channel (611 clauses) and the figure channel (189 derived requirements across 56 built-environment entities). All figures are descriptive counts over the corpus census; every number in the interface is interpolated live from the embedded data blob.

Overview

Corpus dashboard: resolution rate by schema dimension (figures channel), deontic-force distribution by field type, cross-channel validation flow with per-topic match rates, SDA category coverage, polysemy burden, residual-ledger composition and figure↔text entity alignment. Panels deep-link into the detailed views.

Network

The full subject–predicate–object graph: every one of the 189 figure requirements draws an edge. Where the requirement object names another entity (word-boundary mention, resolved at build time), the edge runs entity→entity; standards references share AS-numbered diamond nodes; the remaining requirements run to deduplicated value nodes (squares) carrying their dimensional payload. Node colour encodes the SDA category set, node size the requirement count, a gold ring polysemy, edge colour the relation family, and every edge tooltip states its resolution basis. The structural layer (dashed) renders the 17 canonical relations (28 attestations across design-category options), anchored onto figure entities wherever the name is the same entity — joining the spatial ontology (Dwelling CONTAINS Bedroom …) to the requirements graph; the chip reports nodes, edges and connected components live.

Elements & Clauses

Master–detail browsers for the 56 entities (requirements grouped by relation, dimensions highlighted, checklist export) and the 611 text clauses (modal verbs highlighted, deontic analysis, linked structural triples, residual-ledger cross-links).

Residual Ledger & Audit Trail (W6.1)

The governance-audit layer of the serialisation schema: per-primitive residual entries with rationale and disposition pathway, and one record per figure-derived requirement keyed by SHA-1 record_uid with provenance and cross-channel status equal to the §5.20 cross-validation. Implementation-debt notes declare the derived (not yet materialised) scope of these rows.

Keyboard

16 switch views · / focus search · Esc close

PhD dissertation Chapter 5 — SDA Standards Explorer. W6.1 residual-ledger / audit-trail views per consolidated visualisation plan §2.6, §1.6 C6. Rebuilt 2026-06-10: v2.0 presentation layer; v2.1 added DP-9 object→entity resolution (189/189 S-P-O coverage) + the materialised term→primitive mapping (run 2603221715); v2.2 consumes the root-cause entity-registry-v2 (DP-10): spec-phrase subjects re-typed as concept + quantity spec, the polysemy census corrected to the head-anchored measurement (13 of 56; the former figure of 28 carried wrong-token assignments — see the 2026-06-10 re-audit), and the structural layer canonicalised to 17 relations. Clause-role, mixed-force and cross-channel derivations unchanged from the remediated v1.2 build.