Core Data Model ERD
Last modified by Robert Schaub on 2025/12/18 12:03
Core Data Model ERD
erDiagram
USER ||--o{ CLAIM : creates
CLAIM ||--o{ EVIDENCE : has
CLAIM ||--o{ SCENARIO : has
SCENARIO ||--o{ VERDICT : assessed
EVIDENCE }o--|| SOURCE : from
USER {
uuid id PK
text name
text email
text role "reader|contributor|moderator|admin"
int contributions_count "cached"
timestamp created_at
}
CLAIM {
uuid id PK
uuid user_id FK
text text
decimal confidence "0-1"
jsonb evidence_summary "cached: top 5 evidence"
text_array source_names "cached: for display"
int scenario_count "cached: count"
timestamp cache_updated_at "cache freshness"
timestamp created_at
timestamp updated_at
}
EVIDENCE {
uuid id PK
uuid claim_id FK
uuid source_id FK
text content
decimal relevance "0-1"
text url
timestamp created_at
}
SOURCE {
uuid id PK
text name
text domain
decimal track_record_score "0-1"
int total_citations
timestamp last_updated
timestamp created_at
}
SCENARIO {
uuid id PK
uuid claim_id FK "belongs to claim"
uuid extracted_from "references evidence_id that provided context"
text description
jsonb assumptions
timestamp created_at
timestamp updated_at
}
VERDICT {
uuid id PK
uuid scenario_id FK "assessed scenario"
text likelihood_range "e.g. 0.40-0.65 (uncertain)"
decimal confidence "0-1"
text explanation_summary "verdict reasoning"
text_array uncertainty_factors "factors affecting confidence"
timestamp created_at
timestamp updated_at
}
Core Data Model ERD - Shows primary business entities and their relationships. Claims have Evidence (sources supporting/refuting) and Scenarios (different contexts for evaluation). Each Scenario is assessed by Verdicts (conclusion about the claim in that scenario context). Evidence comes from Sources (with track records). Verdicts track changes through the Edit entity like all other entities. Claims include denormalized cache fields for performance. Most entities created/edited by AKEL automatically. See Audit Trail ERD for edit tracking relationships.