Skip to Content

Data Examples

Last modified by Robert Schaub on 2025/12/18 12:03

Data Examples

The following examples illustrate complete, realistic FactHarbor data objects
across Claims, Scenarios, Evidence, ScenarioEvidenceLinks, Verdicts, and
Re-evaluation behavior.
Each field is annotated with:

  • [A] — fully automatable
  • [M] — AI-draft, human validation needed
  • [H] — human-only
  • [F] — federation metadata (optional)
    These examples conform to:
  • Data Model (Ch. 5)
  • Workflows (Ch. 6)
  • Requirements (Ch. 2)
  • Architecture (Ch. 3)
  • AKEL (Ch. 4)

Example A — “Hydrogen cars are more energy efficient than EVs.”

A clearly empirical, technical, domain-specific claim.

1. Claim (ClaimID: C_H2EV)

FieldValueNotes
ClaimIDC_H2EV
VersionIDv1
Text“Hydrogen cars are more energy efficient than battery electric vehicles (EVs).”[H]
Domainenergy_transport[A]
ClaimTypeliteral[A]
Evaluabilityempirical[A]
RiskTiermedium[A]
ClusterIDCL_EnergyEff[A]
Statusactive

2. Scenario S_H2EV_01 — “Well-to-wheel efficiency, EU grid mix”

FieldValue
ScenarioIDS_H2EV_01
VersionIDv1
ClaimIDC_H2EV
Definitions“Well-to-wheel efficiency” = total chain efficiency. [H]
AssumptionsEU 2020–24 grid mix, electrolysis 69%, compression losses 10%, fuel cell 55%. [M]
ContextBoundaryEurope, 2020–2024 technology[H]
EvaluationMethodComparative WTW energy analysis[A]
RiskTierlow
Statusactive

3. Evidence

3.1 Evidence E1

Peer-reviewed energy-systems paper.

FieldValue
EvidenceIDE_H2EV_Paper1
VersionIDv1
Typescientific_paper
Categoryempirical
Reliabilityhigh
ProvenanceJournal publication + DOI
ExtractionMethodAKEL + human verification
Statusverified

3.2 Evidence E2

Official EU dataset.

FieldValue
EvidenceIDE_H2EV_Dataset
VersionIDv1
Typedataset
Categoryempirical
Reliabilitymedium
ProvenanceEU Energy Stats 2023
ExtractionMethodAPI import
Statusverified
ScenarioEvidenceRelevanceScore / Notes
S_H2EV_01 v1E_H2EV_Paper1 v10.92 [M]
S_H2EV_01 v1E_H2EV_Dataset v10.77 [M]

5. Verdict

5.1 Verdict V_H2EV_01 v1

FieldValue
VerdictIDV_H2EV_01
VersionIDv1
ClaimIDC_H2EV
ScenarioIDS_H2EV_01
EvidenceVersionSet[E_H2EV_Paper1:v1, E_H2EV_Dataset:v1]
LikelihoodRange0.10–0.25
ExplanationSummaryEVs convert grid electricity to motion more efficiently than hydrogen fuel cell vehicles under EU assumptions.
ReasoningChainStep-by-step efficiency-chain comparison
UncertaintyFactorsVariability in grid mix, future electrolysis improvements
Statuscurrent

6. Notes

  • Many fields automatable (claim classification, domain, initial scenario structure, metadata extraction).
  • Definitions and boundaries require human reasoning.

Example B — “Regular cold-water exposure (<14°C) for sufficient months improves health.”

A complex lifestyle/health claim requiring careful scenario boundaries.

7. Claim (C_CW_Health)

FieldValue
Text“Regular cold-water exposure below 14°C for at least 6 months improves health.”
Domainhealth_lifestyle
ClaimTypeliteral
Evaluabilityempirical
RiskTierhigh

8. Scenario S_CW_01 — “Short daily immersions in healthy adults”

FieldValue
ScenarioIDS_CW_01
VersionIDv1
Definitions“Regular exposure” = 3–7×/week, 2–4 minutes
AssumptionsHealthy adults, no cardiovascular risk
ContextBoundary6+ months, 8–14°C
EvaluationMethodHealth outcome comparison
RiskTierhigh

9. Evidence

9.1 Evidence E1 — Dutch cold-shower RCT

FieldValue
Typescientific_paper
Categoryempirical
Reliabilityhigh
ExtractionMethodAKEL + human validation

9.2 Evidence E2 — Meta-analysis on immersion effects

FieldValue
Categoryempirical
Reliabilitymedium
ScenarioEvidenceScore
S_CW_01E10.82
S_CW_01E20.75

11. Verdict

FieldValue
LikelihoodRange0.40–0.65 (uncertain)
ExplanationSummarySome benefits (mood, perceived recovery), but long-term health improvement unclear.
UncertaintyFactorsSmall sample sizes, lifestyle confounds

12. Notes

  • Medical ethics → high human involvement.
  • AKEL helpful for metadata, summaries, and links.

Example C — “Hillary Clinton communicates with Eleanor Roosevelt.”

A belief/metaphorical/non-falsifiable claim.

13. Claim (C_HC_ER)

FieldValue
Text“Hillary Clinton communicates with Eleanor Roosevelt.”
Domainpolitics_private_beliefs
ClaimTypemetaphorical
Evaluabilitynon-falsifiable
RiskTierlow

14. Scenario S_ER_01 — Literal paranormal interpretation

FieldValue
Definitions“Communicate” = literal paranormal communication [H]
AssumptionsParanormal abilities exist [H]
EvaluationMethodNot empirically testable
RiskTierlow
Evaluabilitynon-falsifiable

15. Evidence

Minimal placeholder:

FieldValue
EvidenceIDE_None
Typenone
Categorynone
Reliabilitylow

16. Verdict

LikelihoodRange: undefined
Status: non-evaluable
Reasoning: claim is non-falsifiable.

Example D — “Hillary Clinton is a witch.”

Clearly rhetorical/metaphorical.

17. Claim (C_HC_Witch)

FieldValue
Text“Hillary Clinton is a witch.”
Domainrhetoric
ClaimTyperhetorical
Evaluabilitynon-falsifiable
RiskTierlow

18. Scenario S_Witch_01 — Literal interpretation

FieldValue
DefinitionsSupernatural definition of “witch”
AssumptionsSupernatural powers exist
EvaluationMethodNon-testable
Evaluabilitynon-falsifiable

19. Evidence

None required.

20. Verdict

Likelihood: undefined
Reasoning: rhetorical, not empirical.

Automation Summary Across Examples

21. Fully Automatable [A]

  • Claim normalization 
  • Claim clustering 
  • Evidence metadata extraction 
  • Initial scenario scaffolding 
  • Reliability heuristics 
  • Relevance ranking 
  • Draft verdicts 
  • Trigger detection 

22. Mixed [M]

  • Assumptions 
  • Context boundaries 
  • Relevance scoring 
  • Reasoning chain 
  • Uncertainty factors 

23. Human-only [H]

  • Definitions 
  • Ethical constraints 
  • High-risk scenario approval 
  • Interpretation of meaning 
  • Final verdict approval