Changes for page Workflows

Last modified by Robert Schaub on 2025/12/24 20:34

From 7.1 to 6.1

From version 6.1

edited by Robert Schaub
on 2025/12/14 18:59

Change comment: Imported from XAR

To version 1.1

edited by Robert Schaub
on 2025/12/11 16:36

Change comment: Imported from XAR

Raw
Rendered

Summary

Page properties (1 modified, 0 added, 0 removed)

Details

Page properties

Content

@@ -1,352 +1,343 @@
  = Workflows =
--This page describes the core workflows for content creation, review, and publication in FactHarbor.
++This chapter defines the core workflows used across the FactHarbor system.
--== Overview ==
++Each workflow describes:
--FactHarbor workflows support three publication modes with risk-based review:
++* Purpose
++* Participants
++* Steps
++* Automation vs. manual work
++* (Wherever applicable) linear ASCII flow: a → b → c → d
--* **Mode 1 (Draft)**: Internal only, failed quality gates or pending review
--* **Mode 2 (AI-Generated)**: Public with AI-generated label, passed quality gates
--* **Mode 3 (Human-Reviewed)**: Public with human-reviewed status, highest trust
++Workflows included:
--Workflows vary by **Risk Tier** (A/B/C) and **Content Type** (Claim, Scenario, Evidence, Verdict).
++1. Claim Workflow
++2. Scenario Workflow
++3. Evidence Workflow
++4. Verdict Workflow
++5. Re-evaluation Workflow
++6. Federation Synchronization Workflow
++7. User Role & Review Workflow
++8. AKEL Workflow
++9. Global Trigger Flow
++10. Entity Lifecycle Notes
  ----
--== Claim Submission & Publication Workflow ==
++== Claim Workflow ==
--=== Step 1: Claim Submission ===
++**Purpose:**
++Transform raw text or input material into a normalized, classified, deduplicated, and versioned claim ready for scenario evaluation.
--**Actor**: Contributor or AKEL
++**Participants:**
++* Contributor
++* AKEL
++* Reviewer
--**Actions**:
--* Submit claim text
--* Provide initial sources (optional for human contributors, mandatory for AKEL)
--* System assigns initial AuthorType (Human or AI)
++=== Steps ===
--**Output**: Claim draft created
++**1. Ingestion**
++* User submits text, URL, transcript, or multi-claim content
++* AKEL extracts one or multiple claims
--=== Step 2: AKEL Processing ===
++**2. Normalization**
++* Standardizes wording
++* Reduces ambiguity
++* Flags implicit assumptions
--**Automated Steps**:
--1. Claim extraction and normalization
--2. Classification (domain, type, evaluability)
--3. Risk tier assignment (A/B/C suggested)
--4. Initial scenario generation
--5. Evidence search
--6. **Contradiction search** (mandatory)
--7. Quality gate validation
++**3. Classification (AKEL draft → Reviewer confirm)**
++* ClaimType
++* Domain
++* Evaluability
++* SafetyCategory
--**Output**: Processed claim with risk tier and quality gate results
++**4. Duplicate & Similarity Detection**
++* Embeddings created
++* Similar claims found
++* Reviewer merges, splits, or confirms uniqueness
--=== Step 3: Quality Gate Checkpoint ===
++**5. Version Creation**
++* New ClaimVersion stored
++* Every edit creates a new immutable version
--**Gates Evaluated**:
--* Source quality
--* Contradiction search completion
--* Uncertainty quantification
--* Structural integrity
++**6. Cluster Assignment**
++* AKEL proposes cluster membership
++* Reviewer confirms
--**Outcomes**:
--* **All gates pass** → Proceed to Mode 2 publication (if Tier B or C)
--* **Any gate fails** → Mode 1 (Draft), flag for human review
--* **Tier A** → Mode 2 with warnings + auto-escalate to expert queue
++**7. Scenario Linking (optional)**
++* Existing scenarios connected
++* AKEL may propose new drafts
--=== Step 4: Publication (Risk-Tier Dependent) ===
++**8. Publication**
++* Claim becomes active and visible
--**Tier C (Low Risk)**:
--* **Direct to Mode 2**: AI-generated, public, clearly labeled
--* User can request human review
--* Sampling audit applies
++**Flow:**
++Ingest → Normalize → Classify → Deduplicate → Cluster → Version → Publish
--**Tier B (Medium Risk)**:
--* **Direct to Mode 2**: AI-generated, public, clearly labeled
--* Higher audit sampling rate
--* High-engagement content may auto-escalate
++----
--**Tier A (High Risk)**:
--* **Mode 2 with warnings**: AI-generated, public, prominent disclaimers
--* **Auto-escalated** to expert review queue
--* User warnings displayed
--* Highest audit sampling rate
++== Scenario Workflow ==
--=== Step 5: Human Review (Optional for B/C, Escalated for A) ===
++**Purpose:**
++Define the specific analytic contexts needed to evaluate each claim.
--**Triggers**:
--* User requests review
--* Audit flags issues
--* High engagement (Tier B)
--* Automatic (Tier A)
++**Participants:**
++* Contributor
++* Reviewer
++* Domain Expert
++* AKEL
--**Process**:
--1. Reviewer/Expert examines claim
--2. Validates quality gates
--3. Checks contradiction search results
--4. Assesses risk tier appropriateness
--5. Decision: Approve, Request Changes, or Reject
++=== Steps ===
--**Outcomes**:
--* **Approved** → Mode 3 (Human-Reviewed)
--* **Changes Requested** → Back to contributor or AKEL for revision
--* **Rejected** → Rejected status with reasoning
++**1. Scenario Proposal**
++* Drafted by contributor or generated by AKEL
------
++**2. Completion of Required Fields**
++Must include:
++* Definitions
++* Assumptions
++* ContextBoundary
++* EvaluationMethod
++* SafetyClass
++* VersionMetadata
--== Scenario Creation Workflow ==
++**3. Safety Interception (AKEL)**
++Flags:
++* non-falsifiable structures
++* pseudoscientific assumptions
++* unsafe contexts
--=== Step 1: Scenario Generation ===
++**4. Redundancy & Conflict Check**
++* Similar scenarios merged
++* Contradictions flagged
--**Automated (AKEL)**:
--* Generate scenarios for claim
--* Define boundaries, assumptions, context
--* Identify evaluation methods
++**5. Reviewer Validation**
++Ensures clarity, neutrality, and methodological validity.
--**Manual (Expert/Reviewer)**:
--* Create custom scenarios
--* Refine AKEL-generated scenarios
--* Add domain-specific nuances
++**6. Expert Approval (mandatory for high-risk domains)**
--=== Step 2: Scenario Validation ===
++**7. Version Storage**
++* Each revision = new ScenarioVersion
--**Quality Checks**:
--* Completeness (definitions, boundaries, assumptions clear)
--* Relevance to claim
--* Evaluability
--* No circular logic
++**Flow:**
++Draft → Validate → Safety Check → Review → Expert Approval → Version → Activate
--**Risk Tier Assignment**:
--* Inherits from parent claim
--* Can be overridden by expert if scenario increases/decreases risk
--
--=== Step 3: Scenario Publication ===
--
--**Mode 2 (AI-Generated)**:
--* Tier B/C scenarios can publish immediately
--* Subject to sampling audits
--
--**Mode 1 (Draft)**:
--* Tier A scenarios default to draft
--* Require expert validation for Mode 2 or Mode 3
--
  ----
--== Evidence Evaluation Workflow ==
++== Evidence Workflow ==
--=== Step 1: Evidence Search & Retrieval ===
++**Purpose:**
++Structure, classify, validate, version, and link evidence to scenarios.
--**AKEL Actions**:
--* Search academic databases, reputable media
--* **Mandatory contradiction search** (counter-evidence, reservations)
--* Extract metadata (author, date, publication, methodology)
--* Assess source reliability
++**Participants:**
++* Contributor
++* Reviewer
++* Domain Expert (when needed)
++* AKEL
--**Quality Requirements**:
--* Primary sources preferred
--* Diverse perspectives included
--* Echo chambers flagged
--* Conflicting evidence acknowledged
++=== Steps ===
--=== Step 2: Evidence Summarization ===
++**1. Evidence Submission**
++* File, dataset, URL, or extracted text
--**AKEL Generates**:
--* Summary of evidence
--* Relevance assessment
--* Reliability score
--* Limitations and caveats
--* Conflicting evidence summary
++**2. Metadata Extraction (AKEL)**
++* EvidenceType
++* Category
++* Provenance
++* Study design
++* ExtractionMethod
++* ReliabilityHints
--**Quality Gate**: Structural integrity, source quality
++**3. Relevance Check**
++Reviewer verifies which scenarios the evidence applies to.
--=== Step 3: Evidence Review ===
++**4. Reliability Assessment**
++* AKEL proposes reliability
++* Reviewer confirms
++* Expert review for complex papers
--**Reviewer/Expert Validates**:
--* Accuracy of summaries
--* Appropriateness of sources
--* Completeness of contradiction search
--* Reliability assessments
++**5. ScenarioEvidenceLink Creation**
++Each link stores:
++* relevance score
++* justification
++* evidence version
--**Outcomes**:
--* **Mode 2**: Evidence summaries published as AI-generated
--* **Mode 3**: After human validation
--* **Mode 1**: Failed quality checks or pending expert review
++**6. Versioning**
++* Any update = new EvidenceVersion
++**Flow:**
++Submit → Extract Metadata → Evaluate Relevance → Score Reliability → Link → Version
++
  ----
--== Verdict Generation Workflow ==
++== Verdict Workflow ==
--=== Step 1: Verdict Computation ===
++**Purpose:**
++Generate likelihood estimates per scenario based on evidence and scenario structure.
--**AKEL Computes**:
--* Verdict across scenarios
--* Confidence scores
--* Uncertainty quantification
--* Key assumptions
--* Limitations
++**Participants:**
++* AKEL (drafts)
++* Reviewer
++* Domain Expert
--**Inputs**:
--* Claim text
--* Scenario definitions
--* Evidence assessments
--* Contradiction search results
++=== Steps ===
--=== Step 2: Verdict Validation ===
++**1. Evidence Aggregation**
++Collect relevant evidence versions.
--**Quality Gates**:
--* All four gates apply (source, contradiction, uncertainty, structure)
--* Reasoning chain must be traceable
--* Assumptions must be explicit
++**2. Draft Verdict Generation (AKEL)**
++Outputs:
++* likelihood range
++* uncertainty factors
++* conflict detection
++* sensitivity analysis
--**Risk Tier Check**:
--* Tier A: Always requires expert validation for Mode 3
--* Tier B: Mode 2 allowed, audit sampling
--* Tier C: Mode 2 default
++**3. Reasoning Draft**
++Structured explanation chain generated by AKEL.
--=== Step 3: Verdict Publication ===
++**4. Reviewer Validation**
++Ensures logic, evidence fit, no hallucinations.
--**Mode 2 (AI-Generated Verdict)**:
--* Clear labeling with confidence scores
--* Uncertainty disclosure
--* Links to reasoning trail
--* User can request expert review
++**5. Expert Review**
++Required for:
++* medicine
++* psychology
++* engineering
++* political misinformation
++* controversial or risky domains
--**Mode 3 (Expert-Validated Verdict)**:
--* Human reviewer/expert stamp
--* Additional commentary (optional)
--* Highest trust level
++**6. Verdict Storage**
++* Every update creates a new VerdictVersion
++**Flow:**
++Aggregate → Draft Verdict → Draft Explanation → Review → Expert Approval → Version
++
  ----
--== Audit Workflow ==
++== Re-evaluation Workflow ==
--=== Step 1: Audit Sampling Selection ===
++**Purpose:**
++Keep verdicts current when evidence or scenarios change.
--**Stratified Sampling**:
--* Risk tier priority (A > B > C)
--* Low confidence scores
--* High traffic content
--* Novel topics
--* User flags
++=== Trigger Types ===
--**Sampling Rates** (Recommendations):
--* Tier A: 30-50%
--* Tier B: 10-20%
--* Tier C: 5-10%
++* Evidence updated, disputed, retracted
++* Scenario assumptions changed
++* Claim reclassification
++* AKEL contradiction detection
++* Federation sync
--=== Step 2: Audit Execution ===
++=== Steps ===
--**Auditor Actions**:
--1. Review sampled AI-generated content
--2. Validate quality gates were properly applied
--3. Check contradiction search completeness
--4. Assess reasoning quality
--5. Identify errors or hallucinations
++**1. Trigger Detection**
++Re-evaluation engine receives event.
--**Audit Outcome**:
--* **Pass**: Content remains in Mode 2, logged as validated
--* **Fail**: Content flagged for review, system improvement triggered
++**2. Impact Analysis**
++Find affected:
++* scenarios
++* evidence links
++* verdicts
--=== Step 3: Feedback Loop ===
++**3. AKEL Draft Re-calculation**
++New:
++* likelihood
++* reasoning
++* uncertainty
--**System Improvements**:
--* Failed audits analyzed for patterns
--* AKEL parameters adjusted
--* Quality gates refined
--* Risk tier assignments recalibrated
++**4. Reviewer Validation**
++**5. Expert Review** (high-risk)
++**6. Version Storage**
--**Transparency**:
--* Audit statistics published periodically
--* Patterns shared with community
--* System improvements documented
++**Flow:**
++Trigger → Analyze → Recompute → Review → Expert → Version
  ----
--== Mode Transition Workflow ==
++== Federation Synchronization Workflow ==
--=== Mode 1 → Mode 2 ===
++**Purpose:**
++Exchange structured data between nodes.
--**Requirements**:
--* All quality gates pass
--* Risk tier B or C (or A with warnings)
--* Contradiction search completed
++=== Steps ===
++1. Detect version changes
++1. Build bundle (diff + Merkle tree + signatures)
++1. Push to peers
++1. Validate lineage + hashes
++1. Resolve conflicts (merge or branch)
++1. Optional re-evaluation
--**Trigger**: Automatic upon quality gate validation
++**Flow:**
++Change → Bundle → Push → Validate → Merge/Fork → Update
--=== Mode 2 → Mode 3 ===
++----
--**Requirements**:
--* Human reviewer/expert validation
--* Quality standards confirmed
--* For Tier A: Expert approval required
--* For Tier B/C: Reviewer approval sufficient
++== User Role & Review Workflow ==
--**Trigger**: Human review completion
++**Purpose:**
++Ensure correctness, neutrality, safety, and resistance to manipulation.
--=== Mode 3 → Mode 1 (Demotion) ===
++=== Steps ===
--**Rare - Only if**:
--* New evidence contradicts verdict
--* Error discovered in reasoning
--* Source retraction
++**1. Submission**
++Claim / scenario / evidence / verdict.
--**Process**:
--1. Content flagged for re-evaluation
--2. Moved to draft (Mode 1)
--3. Re-processed through workflow
--4. Reason for demotion documented
++**2. Auto-check (AKEL)**
++Flags unsafe content, contradictions, format issues.
------
++**3. Reviewer Validation**
--== User Actions Across Modes ==
++**4. Expert Validation**
++Required for sensitive domains.
--=== On Mode 1 (Draft) Content ===
++**5. Moderator Oversight**
++Triggered by suspicious behavior.
--**Contributors**:
--* Edit their own drafts
--* Submit for review
++**Flow:**
++Submit → Auto-check → Review → Expert → Moderator (if needed)
--**Reviewers/Experts**:
--* View and comment
--* Request changes
--* Approve for Mode 2 or Mode 3
++----
--=== On Mode 2 (AI-Generated) Content ===
++== AKEL Workflow ==
--**All Users**:
--* Read and use content
--* Request human review
--* Flag for expert attention
--* Provide feedback
++**Purpose:**
++Support extraction, drafting, structuring, and contradiction detection.
--**Reviewers/Experts**:
--* Validate for Mode 3 transition
--* Edit and refine
--* Adjust risk tier if needed
++=== Stages ===
--=== On Mode 3 (Human-Reviewed) Content ===
++**A — Input Understanding:**
++Extraction, normalization, classification.
--**All Users**:
--* Read with highest confidence
--* Still can flag if new evidence emerges
++**B — Scenario Drafting:**
++Definitions, boundaries, assumptions.
--**Reviewers/Experts**:
--* Update if needed
--* Trigger re-evaluation if new evidence
++**C — Evidence Processing:**
++Retrieval, summarization, ranking.
------
++**D — Verdict Drafting:**
++Likelihood, explanations, uncertainties.
--== Diagram References ==
++**E — Safety & Integrity:**
++Contradictions, hallucination detection.
--{{include reference="FactHarbor.Organisation.Diagrams.Claim-Scenario-Lifecycle"/}}
++**F — Human Approval:**
++Reviewer and/or expert.
--{{include reference="FactHarbor.Specification.Diagrams.Manual vs Automated matrix.WebHome"/}}
++**Flow:**
++Input → Drafts → Integrity → Human Approval
  ----
--== Related Pages ==
++== Global Trigger Flow (Cascade) ==
--* [[AKEL (AI Knowledge Extraction Layer)>>FactHarbor.Specification.AI Knowledge Extraction Layer (AKEL).WebHome]]
--* [[Automation>>FactHarbor.Specification.Automation.WebHome]]
--* [[Requirements (Roles)>>FactHarbor.Specification.Requirements.WebHome]]
--* [[Governance>>FactHarbor.Organisation.Governance]]
++Trigger Sources:
++* Claim change
++* Scenario change
++* Evidence change
++* Verdict contradiction
++* Federation update
++* AKEL model improvements
++**Cascade Flow:**
++Trigger → Dependency Graph → Re-evaluation → Updated Verdicts
++

Changes for page Workflows

Summary

Details

Applications

Navigation

Need help?