FAQ Question Pattern Analysis Workflow
Systematic process for generating and validating FAQ content from course materials
flowchart TD
Start(("Course Materials
Assembled")):::startNode
Extract["Extract Concept List
from Learning Graph"]:::autoNode
Analyze["Analyze Concept
Dependencies"]:::autoNode
Search["Search for
Question Patterns"]:::autoNode
Generate["Generate Candidate
Questions (5-10 per concept)"]:::autoNode
QualityCheck{"Quality
Threshold Met?"}:::decisionNode
Flag["Flag for
Human Review"]:::reviewNode
AddDB["Add to FAQ
Database"]:::autoNode
Review["Educator Review
& Edit"]:::reviewNode
GenAnswers["Generate Answers
from Corpus"]:::autoNode
Validate["Validate Answer
Completeness"]:::validNode
Complete{"Answer
Complete?"}:::decisionNode
Revise["Revise Answer"]:::reviewNode
Approve["Approve FAQ Entry"]:::validNode
Update["Update FAQ Index"]:::autoNode
End(("FAQ Published")):::endNode
Start --> Extract
Extract --> Analyze
Analyze --> Search
Search --> Generate
Generate --> QualityCheck
QualityCheck -->|No| Flag
QualityCheck -->|Yes| AddDB
Flag --> Review
Review --> AddDB
AddDB --> GenAnswers
GenAnswers --> Validate
Validate --> Complete
Complete -->|No| Revise
Revise --> GenAnswers
Complete -->|Yes| Approve
Approve --> Update
Update --> End
classDef startNode fill:#667eea,stroke:#333,stroke-width:3px,color:#fff,font-size:16px
classDef autoNode fill:#4facfe,stroke:#333,stroke-width:2px,color:#fff,font-size:16px
classDef decisionNode fill:#ffd166,stroke:#333,stroke-width:2px,color:#333,font-size:16px
classDef reviewNode fill:#f77f00,stroke:#333,stroke-width:2px,color:#fff,font-size:16px
classDef validNode fill:#43aa8b,stroke:#333,stroke-width:2px,color:#fff,font-size:16px
classDef endNode fill:#9b59b6,stroke:#333,stroke-width:3px,color:#fff,font-size:16px
linkStyle default stroke:#999,stroke-width:2px,font-size:14px
Workflow Overview
This systematic workflow uses both automated analysis and human review to generate high-quality FAQ content from course materials including learning graphs, glossaries, and chapter content.
Workflow Phases
Phase 1: Automated Analysis
- Extract Concept List - Parse learning graph to enumerate all concepts
- Analyze Dependencies - Identify complex concepts (high in-degree) and foundational concepts (zero dependencies)
- Search Patterns - Scan corpus for existing questions and common interrogative structures
- Generate Questions - Use Claude API to create 5-10 questions per concept across categories:
- Definitional: "What is..."
- Procedural: "How do I..."
- Troubleshooting: "Why doesn't..."
- Comparative: "What's the difference between..."
Phase 2: Quality Check
Questions must meet criteria:
- ✓ Non-redundant (unique from existing FAQs)
- ✓ Answerable from course content (no external knowledge required)
- ✓ Aligned with reading level
- ✓ Diverse across question categories
Passing rate: Typically 60-70% of auto-generated questions pass quality check
Phase 3: Human Review
Subject matter expert reviews flagged questions and:
- Edits for clarity and accuracy
- Ensures pedagogical appropriateness
- Removes questions outside course scope
- Adds domain-specific nuance
Phase 4: Answer Generation & Validation
- Generate Answers - Claude retrieves relevant passages from course content and synthesizes comprehensive answers with chapter citations
- Validate Completeness - Check that answers directly address questions, stay within course scope, and match reading level
- Revise if Needed - Educator supplements or rewrites answers; may identify content gaps requiring new chapter sections
- Approve Entry - Finalized FAQ question-answer pair added to /docs/faq.md
Phase 5: Publication
Update FAQ Index - Add search keywords, concept tags, and navigation links; integrate into MkDocs site
Analytics Tracking - Monitor which questions receive most views to identify student needs
Quality Metrics
Target: 100-200 FAQ entries for a complete course
Coverage: Every concept should have at least 1 FAQ question
Human intervention: 30-40% of questions require educator review
Answer completeness: 95%+ of answers should be self-contained (no external resources needed)
Integration Benefits
- Chatbot Ready: Structured FAQ database enables AI chatbot integration
- Search Optimized: FAQ content improves site search results
- Student Support: Reduces instructor questions by addressing common issues proactively
- Content Validation: FAQ generation process often reveals gaps in course content