# Personal Knowledge Graph
A personal knowledge graph represents a person's people, accounts, devices, places, events, organizations, claims, and source artifacts as typed entities connected by explicit relationships and provenance.
## Notebook precursor
[[Scanned_20260730-1913]] performs this work manually. Contact slips, residence dates, device tasks, account lineages, receipts, routes, company lists, diagrams, and conceptual maps become cross-referencable nodes. The notebook's weakness is not absence of graph thinking but the absence of machine-readable edge types, confidence, source identifiers, dates, and privacy classes.
## Required archival fields
- canonical entity and aliases;
- exact source notebook and page;
- visible evidence versus correction or later interpretation;
- relationship type and direction;
- date or date range;
- confidence and unresolved alternatives;
- sensitivity class, with secrets stored outside the public graph.
## Relationships
[[Control-Plane Continuity]] · [[Identity and Account Continuity]] · [[Documentary Evidence Continuity]] · [[Personal Digital Twin]] · [[Index - Pattern Ledger]]
## Source
- [[Scanned_20260730-1913]], notebook-level synthesis.
## Scanned_20260730-1845 overlay
PDF page 27 of [[Scanned_20260730-1845]] explicitly asks for a visual graph of every company and its historical data. The notebook supplies a broader schema than the earlier paper continuity ledger: companies and acquisitions, domains and accounts, certificate issuers, official credentials, devices, permissions, travel/insurance records, legal repositories, and semantic goals.
The decisive addition is time-aware [[Data Provenance]]. A name or platform must be linked to former names, owners, source records, dates, and confidence rather than treated as one timeless node.