Framework: data model, grouping rules, and representation
The kihubstat framework defines a neutral observational model for recording and organising activity records. The model separates primary event objects from associative layers. Primary events contain canonical fields such as source identifier, timestamp, actor role, context tags, and an unstructured observation text. Each primary event additionally retains its raw payload to preserve original context. Associative layers hold annotations, curated groupings, and link objects that reference primary events without altering them. This separation maintains the fidelity of original observations while enabling structured layers for review and commentary. The design supports deterministic grouping based on shared keys, temporal proximity, or source metadata, as well as curator-defined collections for targeted archival inspection. Interfaces emphasise clear provenance information and audit trails that list actor, rationale, and timestamps for every associative operation.
Data structures and grouping model
Records are organised into a canonical schema designed for archival clarity. A primary event comprises: an immutable identifier, a recorded timestamp with timezone context, a source reference, an actor descriptor indicating role and affiliation, contextual tags drawn from controlled vocabularies, and an unstructured observation body that preserves the observer's notes. A raw payload field stores the original submission to support full-fidelity audits. Grouping is expressed through separate collection objects. Deterministic grouping uses matching keys such as process identifiers, correlation tokens, or temporal windows. Curator-defined collections allow observers to assemble thematic sets for review. Link objects are first-class items that reference events and store metadata: author, rationale, and timestamp. Because associative objects are stored independently, multiple overlapping groupings can coexist, enabling comparative review without altering original records. Indexing supports faceted retrieval across canonical fields and controlled tags to make archival inspection methodical and reproducible.
Normalization and audit
Normalization maps variant field names from diverse sources to the canonical schema while preserving raw inputs. Every transformation step is recorded in the audit trail. Audits capture actor identity, time, and descriptive rationale for each associative action or normalization rule application. This produces a verifiable chain of custody for each record and its associative layers.
Visualization paradigms and interfaces
Visualization in kihubstat focuses on editorial clarity and reference utility rather than metric-oriented dashboards. Chronological views present ordered streams that retain contextual metadata and allow inline annotation. Relational maps render events as nodes with typed edges for link objects; layout options include compact radial views and layered orthogonal maps to support inspection at varying scopes. Reference summaries compile curated excerpts and annotated notes to form durable archival pages that combine primary excerpts with associative commentary. Visual components follow an archival aesthetic: subdued palettes, generous typographic hierarchy, and asymmetrical grids that juxtapose text blocks, citation rails, and schematic diagrams. Interfaces provide adjustable focus controls to filter associative layers, collapse or expand groupings, and reveal provenance metadata on demand. Export routines produce structured extracts that include both canonical records and raw payloads as separate bundles suitable for archival transfer.
Access control and retention considerations
Access is role-based and oriented around review permissions. Primary records are immutable by design; annotations and groupings are separate objects subject to access controls. Audit logs record all editorial actions, including author, timestamp, and rationale. Retention rules are configurable per environment and documented alongside export policies. Export bundles preserve normalized records and raw payloads as distinct archives to maintain provenance. The framework supports queryable audit trails to enable methodical archival reviews and compliance-oriented inspections while preserving descriptive integrity of original observations.