Skip to Content

Changes for page Automation

Last modified by Robert Schaub on 2025/12/22 14:38
From version 1.2
edited by Robert Schaub
on 2025/12/22 14:38
Change comment: Update document after refactoring.
To version 1.5
edited by Robert Schaub
on 2025/12/22 14:38
Change comment: Renamed back-links.

Summary

Details

Page properties
Content
... ... @@ -1,21 +1,31 @@
1 1  = Automation =
2 +
2 2  **How FactHarbor scales through automated claim evaluation.**
4 +
3 3  == 1. Automation Philosophy ==
6 +
4 4  FactHarbor is **automation-first**: AKEL (AI Knowledge Extraction Layer) makes all content decisions. Humans monitor system performance and improve algorithms.
5 5  **Why automation:**
9 +
6 6  * **Scale**: Can process millions of claims
7 7  * **Consistency**: Same evaluation criteria applied uniformly
8 8  * **Transparency**: Algorithms are auditable
9 9  * **Speed**: Results in <20 seconds typically
10 10  See [[Automation Philosophy>>Test.FactHarbor.Organisation.Automation-Philosophy]] for detailed principles.
15 +
11 11  == 2. Claim Processing Flow ==
17 +
12 12  === 2.1 User Submits Claim ===
19 +
13 13  * User provides claim text + source URLs
14 14  * System validates format
15 15  * Assigns processing ID
16 16  * Queues for AKEL processing
24 +
17 17  === 2.2 AKEL Processing ===
26 +
18 18  **AKEL automatically:**
28 +
19 19  1. Parses claim into testable components
20 20  2. Extracts evidence from sources
21 21  3. Scores source credibility
... ... @@ -25,9 +25,12 @@
25 25  7. Publishes result
26 26  **Processing time**: Typically <20 seconds
27 27  **No human approval required** - publication is automatic
38 +
28 28  === 2.3 Publication States ===
40 +
29 29  **Processing**: AKEL working on claim (not visible to public)
30 30  **Published**: AKEL completed evaluation (public)
43 +
31 31  * Verdict displayed with confidence score
32 32  * Evidence and sources shown
33 33  * Risk tier indicated
... ... @@ -72,33 +72,45 @@
72 72  * **Human review queue** for low-confidence verdicts
73 73  * **Independent claim processing** - errors in one claim don't cascade to others == 3. Risk Tiers ==
74 74  Risk tiers classify claims by potential impact and guide audit sampling rates.
88 +
75 75  === 3.1 Tier A (High Risk) ===
90 +
76 76  **Domains**: Medical, legal, elections, safety, security
77 77  **Characteristics**:
93 +
78 78  * High potential for harm if incorrect
79 79  * Complex specialized knowledge required
80 80  * Often subject to regulation
81 81  **Publication**: AKEL publishes automatically with prominent risk warning
82 82  **Audit rate**: Higher sampling recommended
99 +
83 83  === 3.2 Tier B (Medium Risk) ===
101 +
84 84  **Domains**: Complex policy, science, causality claims
85 85  **Characteristics**:
104 +
86 86  * Moderate potential impact
87 87  * Requires careful evidence evaluation
88 88  * Multiple valid interpretations possible
89 89  **Publication**: AKEL publishes automatically with standard risk label
90 90  **Audit rate**: Moderate sampling recommended
110 +
91 91  === 3.3 Tier C (Low Risk) ===
112 +
92 92  **Domains**: Definitions, established facts, historical data
93 93  **Characteristics**:
115 +
94 94  * Low potential for harm
95 95  * Well-documented information
96 96  * Clear right/wrong answers typically
97 97  **Publication**: AKEL publishes by default
98 98  **Audit rate**: Lower sampling recommended
121 +
99 99  == 4. Quality Gates ==
123 +
100 100  AKEL applies quality gates before publication. If any fail, claim is **flagged** (not blocked - still published).
101 101  **Quality gates**:
126 +
102 102  * Sufficient evidence extracted (≥2 sources)
103 103  * Sources meet minimum credibility threshold
104 104  * Confidence score calculable
... ... @@ -105,19 +105,21 @@
105 105  * No detected manipulation patterns
106 106  * Claim parseable into testable form
107 107  **Failed gates**: Claim published with flag for moderator review
133 +
108 108  == 5. Automation Levels ==
109 -{{include reference="Test.FactHarbor.Specification.Diagrams.Automation Level.WebHome"/}}
135 +
136 +{{include reference="Test.FactHarbor pre13 V0\.9\.70.Specification.Diagrams.Automation Level.WebHome"/}}
110 110  FactHarbor progresses through automation maturity levels:
111 111  **Release 0.5** (Proof-of-Concept): Tier C only, human review required
112 112  **Release 1.0** (Initial): Tier B/C auto-published, Tier A flagged for review
113 113  **Release 2.0** (Mature): All tiers auto-published with risk labels, sampling audits
114 -See [[Automation Roadmap>>Test.FactHarbor.Specification.Diagrams.Automation Roadmap.WebHome]] for detailed progression. == 5.5 Automation Roadmap == {{include reference="Test.FactHarbor.Specification.Diagrams.Automation Roadmap.WebHome"/}} == 6. Human Role ==
141 +See [[Automation Roadmap>>Test.FactHarbor pre13 V0\.9\.70.Specification.Diagrams.Automation Roadmap.WebHome]] for detailed progression. == 5.5 Automation Roadmap == {{include reference="Test.FactHarbor pre13 V0\.9\.70.Specification.Diagrams.Automation Roadmap.WebHome"/}} == 6. Human Role ==
115 115  Humans do NOT review content for approval. Instead:
116 116  **Monitoring**: Watch aggregate performance metrics
117 117  **Improvement**: Fix algorithms when patterns show issues
118 118  **Exception handling**: Review AKEL-flagged items
119 119  **Governance**: Set policies AKEL applies
120 -See [[Contributor Processes>>Test.FactHarbor.Organisation.Contributor-Processes]] for how to improve the system. == 6.5 Manual vs Automated Matrix == {{include reference="Test.FactHarbor.Specification.Diagrams.Manual vs Automated matrix.WebHome"/}} == 7. Moderation ==
147 +See [[Contributor Processes>>Test.FactHarbor.Organisation.Contributor-Processes]] for how to improve the system. == 6.5 Manual vs Automated Matrix == {{include reference="Test.FactHarbor pre13 V0\.9\.70.Specification.Diagrams.Manual vs Automated matrix.WebHome"/}} == 7. Moderation ==
121 121  Moderators handle items AKEL flags:
122 122  **Abuse detection**: Spam, manipulation, harassment
123 123  **Safety issues**: Content that could cause immediate harm
... ... @@ -125,7 +125,9 @@
125 125  **Action**: May temporarily hide content, ban users, or propose algorithm improvements
126 126  **Does NOT**: Routinely review claims or override verdicts
127 127  See [[Organisational Model>>Test.FactHarbor.Organisation.Organisational-Model]] for moderator role details.
155 +
128 128  == 8. Continuous Improvement ==
157 +
129 129  **Performance monitoring**: Track AKEL accuracy, speed, coverage
130 130  **Issue identification**: Find systematic errors from metrics
131 131  **Algorithm updates**: Deploy improvements to fix patterns
... ... @@ -132,15 +132,21 @@
132 132  **A/B testing**: Validate changes before full rollout
133 133  **Retrospectives**: Learn from failures systematically
134 134  See [[Continuous Improvement>>Test.FactHarbor.Organisation.How-We-Work-Together.Continuous-Improvement]] for improvement cycle.
164 +
135 135  == 9. Scalability ==
166 +
136 136  Automation enables FactHarbor to scale:
168 +
137 137  * **Millions of claims** processable
138 138  * **Consistent quality** at any volume
139 139  * **Cost efficiency** through automation
140 140  * **Rapid iteration** on algorithms
141 141  Without automation: Human review doesn't scale, creates bottlenecks, introduces inconsistency.
174 +
142 142  == 10. Transparency ==
176 +
143 143  All automation is transparent:
178 +
144 144  * **Algorithm parameters** documented
145 145  * **Evaluation criteria** public
146 146  * **Source scoring rules** explicit