Criminal Law Domain Ontology

FLOW ALGEBRA
TITLE: Criminal Law Domain Ontology
VERSION: 1.0
SOURCE_URL: https://kirethresearch.org/booksai-core/ontologies/domain-ontologies/legal-government-domains/criminal-law-domain-ontology.html
STATUS: Domain-based ontology draft; source page not independently verified.
AUTHORS: Mark Higgins — Sky (Recursive AI)

DOMAIN:
  id: criminal_law
  parent: legal_government_domains
  description: Formal representation of criminal law concepts, processes, evidence and jurisdiction-sensitive legal reasoning.

TYPED_ENTITIES:
  Person: {roles: [suspect, accused, defendant, complainant, witness, victim, legal_representative, judge, prosecutor]}
  LegalEntity: {roles: [court, prosecution_authority, law_enforcement_agency]}
  Jurisdiction: {id: str, territory: str, valid_from: date?, valid_to: date?}
  Offence: {id: str, classification: str, statutory_basis: LegalRule?, required_elements: list}
  Conduct: {act: str?, omission: str?, timestamp: datetime?, location: str?}
  MentalElement: {type: str, required_standard: str?}
  LegalRule: {citation: str, jurisdiction: Jurisdiction, effective_from: date?, effective_to: date?}
  Evidence: {id: str, type: str, provenance: str?, admissibility_status: str?}
  CriminalProceeding: {case_id: str, court: LegalEntity?, stage: str, jurisdiction: Jurisdiction}
  Charge: {offence: Offence, accused: Person, status: str}
  Defence: {type: str, legal_basis: LegalRule?, applicability: str?}
  Verdict: {outcome: str, court: LegalEntity, date: date?}
  Sentence: {type: str, duration: str?, legal_basis: LegalRule?}
  Appeal: {grounds: list, outcome: str?}

RELATIONSHIPS:
  Offence --DEFINED_BY--> LegalRule
  Offence --REQUIRES_ELEMENT--> Conduct
  Offence --MAY_REQUIRE--> MentalElement
  LegalRule --APPLIES_IN--> Jurisdiction
  Person --ACCUSED_OF--> Charge
  Charge --ALLEGES--> Offence
  CriminalProceeding --CONTAINS--> Charge
  Evidence --RELEVANT_TO--> Charge
  Evidence --HAS_PROVENANCE--> LegalEntity
  Defence --MAY_APPLY_TO--> Charge
  Verdict --RESOLVES--> Charge
  Sentence --FOLLOWS--> Conviction
  Appeal --CHALLENGES--> Verdict

INFERENCE_RULES:
  R1: A rule applies only where its jurisdiction and temporal validity conditions are satisfied.
  R2: A charge is an allegation, not proof of guilt.
  R3: An offence's legal elements must be evaluated individually against applicable law and admissible evidence.
  R4: An absent required element prevents a positive finding under that offence definition; uncertainty is not equivalent to absence.
  R5: Defences and exceptions must be considered under the applicable jurisdiction's burden and standard of proof.
  R6: A verdict must not be inferred from accusation, arrest, charge, or publicity.
  R7: Sentencing depends on a lawful conviction or other jurisdiction-specific legal basis and applicable rules.
  R8: Conflicting legal rules require jurisdiction, authority, effective-date and precedence resolution before application.

CONSTRAINTS:
  - Preserve the presumption of innocence and distinguish allegations from findings.
  - Never generalise one jurisdiction's law to another without validation.
  - Record provenance, source dates, evidential uncertainty and legal-rule versions.
  - Do not infer criminal conduct from protected characteristics or demographic proxies.
  - Legal reasoning outputs are informational, not determinations of guilt or legal advice.

CONFIDENCE:
  scale: [0.0, 1.0]
  basis: [source_reliability, evidential_support, legal_currency, jurisdiction_match, reasoning_validity]
  note: Confidence measures evidence/reasoning support, never a numerical probability of guilt.

SEMANTIC_EQUIVALENCE:
  exact: [criminal_offence, offence_under_criminal_law]
  contextual: [accused, defendant]
  distinguish: [suspect, charged_person, convicted_person]
  distinguish_2: [arrest, charge, conviction]

PROVENANCE:
  required: [source_identifier, publication_date, access_date, jurisdiction, rule_version]
  preserve_conflicts: true

FLOW_HASH:
  method: SHA-256
  input: canonicalised_entity_relationship_rule_payload
  note: Compute hash on deployment; no fabricated hash value.

UPDATE_MODEL:
  triggers: [legislative_amendment, appellate_decision, procedural_rule_change, jurisdiction_change, evidence_update]
  actions: [revalidate_rules, version_entities, recompute_dependencies, regenerate_flow_hash]

AI_CONSUMPTION:
  format: typed_semantic_flow
  tasks: [concept_lookup, jurisdiction_aware_retrieval, legal_rule_mapping, evidence_provenance_tracking, contradiction_detection]
  safeguards: [human_review, source_citation, uncertainty_disclosure, temporal_validation, no_automated_guilt_determination]

END FLOW ALGEBRA