Overview#
Investigations, provenance review, entity resolution, and intelligence analysis all depend on one basic question: what is connected to what? The Relationship Graph Layer gives analysts a fast, governed way to expand from one entity to its neighbours, find paths between entities, retrieve provenance chains, and support similarity search without exposing the storage implementation behind those results.
The module presents relationship traversal as a platform capability rather than a database feature. It keeps relationship access organisation-scoped, respects soft-deleted and restricted records, supports depth-limited traversal, and provides the graph foundation used by case workspaces, profile intelligence, evidence provenance, and link analysis.
Last Reviewed: 2026-06-26 Last Updated: 2026-06-26
Key Features#
- Depth-Limited Traversal: Analysts can expand a graph around an entity to a controlled depth, reducing noise and preventing runaway exploration.
- Shortest Path Review: Find whether two entities are connected and show the relationship chain that explains the connection.
- Provenance Chain Retrieval: Trace how an entity, evidence item, or derived record was generated, transformed, and linked over time.
- Entity Similarity Search: Support deduplication, link suggestion, and semantic discovery by comparing entity representations.
- Organisation-Scoped Isolation: Relationship results are confined to the authenticated organisation and authorised case scope.
- Soft-Delete Awareness: Deleted or withdrawn relationships are excluded from ordinary analysis while preserving audit and provenance evidence.
- Open Standards Alignment: Provenance and relationship outputs align with standard provenance and linked-data models.
Use Cases#
- Investigation Link Analysis: An analyst expands from a person to related vehicles, addresses, accounts, cases, and evidence records.
- Case Deconfliction: Teams check whether two investigations share entities before planning operational activity.
- Evidence Lineage Review: Reviewers trace how an OCR text record, redacted document, or exported package relates back to source evidence.
- Entity Deduplication: Similar profiles are surfaced for human review before they are merged or linked.
- Operational Briefing: Commanders receive a concise relationship map showing the entities that matter for a current operation.
Integration#
The Relationship Graph Layer connects to profile intelligence, case management, evidence provenance, alert correlation, search, entity resolution, and briefing exports. Access is governed by the same organisation, role, clearance, and case visibility controls used by the source records, so graph exploration cannot bypass normal data boundaries.
Open Standards#
- W3C PROV-DM: Provenance relationships follow the W3C Provenance Data Model for generation, attribution, association, derivation, and communication.
- W3C PROV-JSON: Provenance chains can be represented in a standards-based JSON form for audit and exchange.
- JSON-LD 1.1: Linked-data exports can use JSON-LD contexts where partner systems need semantic relationship exchange.
- ISO 8601: Relationship, provenance, and audit timestamps use standard date-time formatting.
- OAuth 2.0 and JWT Bearer Token: Token-based authentication protects graph traversal, provenance, and similarity workflows.
All relationship traversal and similarity operations are exposed as governed platform capabilities with no public dependency on a particular storage engine.