Skip to Content

Wiki source code of Data Examples

Last modified by Robert Schaub on 2025/12/18 12:03
Show last authors
1 = Data Examples =
2 The following examples illustrate complete, realistic FactHarbor data objects
3 across Claims, Scenarios, Evidence, ScenarioEvidenceLinks, Verdicts, and
4 Re-evaluation behavior.
5 Each field is annotated with:
6 * **[A]** — fully automatable
7 * **[M]** — AI-draft, human validation needed
8 * **[H]** — human-only
9 * **[F]** — federation metadata (optional)
10 These examples conform to:
11 * Data Model (Ch. 5)
12 * Workflows (Ch. 6)
13 * Requirements (Ch. 2)
14 * Architecture (Ch. 3)
15 * AKEL (Ch. 4)
16 = Example A — “Hydrogen cars are more energy efficient than EVs.” =
17 A clearly empirical, technical, domain-specific claim.
18 == 1. Claim (ClaimID: C_H2EV) ==
19 (% style="width:100%" %)
20 |=Field|=Value|=Notes
21 |ClaimID|C_H2EV|
22 |VersionID|v1|
23 |Text|“Hydrogen cars are more energy efficient than battery electric vehicles (EVs).”|[H]
24 |Domain|energy_transport|[A]
25 |ClaimType|literal|[A]
26 |Evaluability|empirical|[A]
27 |RiskTier|medium|[A]
28 |ClusterID|CL_EnergyEff|[A]
29 |Status|active|
30 == 2. Scenario S_H2EV_01 — “Well-to-wheel efficiency, EU grid mix” ==
31 (% style="width:100%" %)
32 |=Field|=Value
33 |ScenarioID|S_H2EV_01
34 |VersionID|v1
35 |ClaimID|C_H2EV
36 |Definitions|//“Well-to-wheel efficiency” = total chain efficiency.// [H]
37 |Assumptions|EU 2020–24 grid mix, electrolysis 69%, compression losses 10%, fuel cell 55%. [M]
38 |ContextBoundary|Europe, 2020–2024 technology|[H]
39 |EvaluationMethod|Comparative WTW energy analysis|[A]
40 |RiskTier|low
41 |Status|active|
42 == 3. Evidence ==
43 === 3.1 Evidence E1 ===
44 Peer-reviewed energy-systems paper.
45 (% style="width:100%" %)
46 |=Field|=Value
47 |EvidenceID|E_H2EV_Paper1
48 |VersionID|v1
49 |Type|scientific_paper
50 |Category|empirical
51 |Reliability|high
52 |Provenance|Journal publication + DOI
53 |ExtractionMethod|AKEL + human verification
54 |Status|verified|
55 === 3.2 Evidence E2 ===
56 Official EU dataset.
57 (% style="width:100%" %)
58 |=Field|=Value
59 |EvidenceID|E_H2EV_Dataset
60 |VersionID|v1
61 |Type|dataset
62 |Category|empirical
63 |Reliability|medium
64 |Provenance|EU Energy Stats 2023
65 |ExtractionMethod|API import
66 |Status|verified|
67 == 4. ScenarioEvidenceLinks ==
68 (% style="width:100%" %)
69 |=(% style="width:30%" %)Scenario|=(% style="width:40%" %)Evidence|=(% style="width:30%" %)RelevanceScore / Notes
70 |S_H2EV_01 v1|E_H2EV_Paper1 v1|0.92 [M]
71 |S_H2EV_01 v1|E_H2EV_Dataset v1|0.77 [M]
72 == 5. Verdict ==
73 === 5.1 Verdict V_H2EV_01 v1 ===
74 (% style="width:100%" %)
75 |=Field|=Value
76 |VerdictID|V_H2EV_01
77 |VersionID|v1
78 |ClaimID|C_H2EV
79 |ScenarioID|S_H2EV_01
80 |EvidenceVersionSet|[E_H2EV_Paper1:v1, E_H2EV_Dataset:v1]
81 |LikelihoodRange|0.10–0.25|
82 |ExplanationSummary|EVs convert grid electricity to motion more efficiently than hydrogen fuel cell vehicles under EU assumptions.|
83 |ReasoningChain|Step-by-step efficiency-chain comparison|
84 |UncertaintyFactors|Variability in grid mix, future electrolysis improvements|
85 |Status|current|
86 == 6. Notes ==
87 * Many fields automatable (claim classification, domain, initial scenario structure, metadata extraction).
88 * Definitions and boundaries require human reasoning.
89 = Example B — “Regular cold-water exposure (<14°C) for sufficient months improves health.” =
90 A complex lifestyle/health claim requiring careful scenario boundaries.
91 == 7. Claim (C_CW_Health) ==
92 (% style="width:100%" %)
93 |=Field|=Value
94 |Text|“Regular cold-water exposure below 14°C for at least 6 months improves health.”|
95 |Domain|health_lifestyle|
96 |ClaimType|literal|
97 |Evaluability|empirical|
98 |RiskTier|high|
99 == 8. Scenario S_CW_01 — “Short daily immersions in healthy adults” ==
100 (% style="width:100%" %)
101 |=Field|=Value
102 |ScenarioID|S_CW_01
103 |VersionID|v1
104 |Definitions|“Regular exposure” = 3–7×/week, 2–4 minutes|
105 |Assumptions|Healthy adults, no cardiovascular risk|
106 |ContextBoundary|6+ months, 8–14°C|
107 |EvaluationMethod|Health outcome comparison|
108 |RiskTier|high|
109 == 9. Evidence ==
110 === 9.1 Evidence E1 — Dutch cold-shower RCT ===
111 (% style="width:100%" %)
112 |=Field|=Value
113 |Type|scientific_paper|
114 |Category|empirical|
115 |Reliability|high|
116 |ExtractionMethod|AKEL + human validation|
117 === 9.2 Evidence E2 — Meta-analysis on immersion effects ===
118 (% style="width:100%" %)
119 |=Field|=Value
120 |Category|empirical|
121 |Reliability|medium|
122 == 10. ScenarioEvidenceLinks ==
123 (% style="width:100%" %)
124 |=Scenario|=Evidence|=Score
125 |S_CW_01|E1|0.82
126 |S_CW_01|E2|0.75
127 == 11. Verdict ==
128 (% style="width:100%" %)
129 |=Field|=Value
130 |LikelihoodRange|0.40–0.65 (uncertain)|
131 |ExplanationSummary|Some benefits (mood, perceived recovery), but long-term health improvement unclear.|
132 |UncertaintyFactors|Small sample sizes, lifestyle confounds|
133 == 12. Notes ==
134 * Medical ethics → high human involvement.
135 * AKEL helpful for metadata, summaries, and links.
136 = Example C — “Hillary Clinton communicates with Eleanor Roosevelt.” =
137 A belief/metaphorical/non-falsifiable claim.
138 == 13. Claim (C_HC_ER) ==
139 (% style="width:100%" %)
140 |=Field|=Value
141 |Text|“Hillary Clinton communicates with Eleanor Roosevelt.”|
142 |Domain|politics_private_beliefs|
143 |ClaimType|metaphorical|
144 |Evaluability|non-falsifiable|
145 |RiskTier|low|
146 == 14. Scenario S_ER_01 — Literal paranormal interpretation ==
147 (% style="width:100%" %)
148 |=Field|=Value
149 |Definitions|“Communicate” = literal paranormal communication [H]|
150 |Assumptions|Paranormal abilities exist [H]|
151 |EvaluationMethod|Not empirically testable|
152 |RiskTier|low|
153 |Evaluability|non-falsifiable|
154 == 15. Evidence ==
155 Minimal placeholder:
156 (% style="width:100%" %)
157 |=Field|=Value
158 |EvidenceID|E_None|
159 |Type|none|
160 |Category|none|
161 |Reliability|low|
162 == 16. Verdict ==
163 LikelihoodRange: **undefined**
164 Status: **non-evaluable**
165 Reasoning: claim is non-falsifiable.
166 = Example D — “Hillary Clinton is a witch.” =
167 Clearly rhetorical/metaphorical.
168 == 17. Claim (C_HC_Witch) ==
169 (% style="width:100%" %)
170 |=Field|=Value
171 |Text|“Hillary Clinton is a witch.”|
172 |Domain|rhetoric|
173 |ClaimType|rhetorical|
174 |Evaluability|non-falsifiable|
175 |RiskTier|low|
176 == 18. Scenario S_Witch_01 — Literal interpretation ==
177 (% style="width:100%" %)
178 |=Field|=Value
179 |Definitions|Supernatural definition of “witch”|
180 |Assumptions|Supernatural powers exist|
181 |EvaluationMethod|Non-testable|
182 |Evaluability|non-falsifiable|
183 == 19. Evidence ==
184 None required.
185 == 20. Verdict ==
186 Likelihood: **undefined**
187 Reasoning: rhetorical, not empirical.
188 = Automation Summary Across Examples =
189 == 21. Fully Automatable [A] ==
190 * Claim normalization
191 * Claim clustering
192 * Evidence metadata extraction
193 * Initial scenario scaffolding
194 * Reliability heuristics
195 * Relevance ranking
196 * Draft verdicts
197 * Trigger detection
198 == 22. Mixed [M] ==
199 * Assumptions
200 * Context boundaries
201 * Relevance scoring
202 * Reasoning chain
203 * Uncertainty factors
204 == 23. Human-only [H] ==
205 * Definitions
206 * Ethical constraints
207 * High-risk scenario approval
208 * Interpretation of meaning
209 * Final verdict approval