1113 Query Framework 99cbd4d4

QUEST Matrix: Systematic Query Generation

A framework for generating comprehensive search queries that cover all angles of a topic.

Overview

QUEST = Questions + Universes + Expansions + Scopes + Types

Goal: Generate 15-25 queries that ensure no major angle is missed.

Step 1: Questions (5W1H Decomposition)

For any topic, decompose into fundamental questions:

Template

## 5W1H for "[TOPIC]"

- **Who:** [List actors, experts, affected parties]
- **What:** [Define, list components/variations]
- **When:** [Timeline, milestones, current state]
- **Where:** [Contexts, platforms, industries, regions]
- **Why:** [Drivers, motivations, benefits sought]
- **How:** [Mechanisms, processes, implementation methods]

Step 2: Universes (Stakeholder Perspectives)

Different stakeholders care about different aspects. Generate queries from each viewpoint:

Template

## Stakeholder Queries for "[TOPIC]"

| Stakeholder | Question They'd Ask | Search Query |
|-------------|---------------------|--------------|
| End Users | "Is this worth it for me?" | "[topic] user reviews pros cons" |
| Providers | "How do I implement this well?" | "[topic] best practices implementation" |
| Regulators | "What are the risks?" | "[topic] risks compliance" |
| Competitors | "How does this compare?" | "[topic] vs alternatives comparison" |
| Experts | "What does research say?" | "[topic] research study findings" |
| Critics | "What could go wrong?" | "[topic] failures problems criticism" |

Step 3: Expansions (Semantic Breadth)

Expand terminology to catch content using different words:

Template

## Semantic Expansion for "[TOPIC]"

| Core Term | Synonyms | Related Terms | Domain Jargon |
|-----------|----------|---------------|---------------|
| [Main term] | [List 2-3] | [List 2-3] | [List 2-3] |
| [Sub-term 1] | [List] | [List] | [List] |
| [Sub-term 2] | [List] | [List] | [List] |

Step 4: Scopes (Boundary Exploration)

Explore different boundaries and contexts:

Template

## Scope Variations for "[TOPIC]"

**Geographic:**
- Local: "[topic] [country/region]"
- Global: "[topic] global trends worldwide"

**Temporal:**
- Historical: "[topic] history evolution"
- Current: "[topic] 2025 state of the art"
- Future: "[topic] future trends predictions"

**Scale:**
- Individual: "[topic] for individuals personal"
- Organization: "[topic] enterprise business"
- Industry: "[topic] industry impact"

Step 5: Types (Source Diversity)

Target different source types for balanced information:

Template

## Source Type Queries for "[TOPIC]"

| Type | Query |
|------|-------|
| Academic | "[topic] research study peer-reviewed" |
| Industry | "[topic] industry report analysis" |
| News | "[topic] news latest 2025" |
| Official | "[topic] official guidelines" |
| User | "[topic] reddit user experience" |
| Expert | "[topic] expert interview opinion" |

Complete QUEST Workflow

Phase 2: PLAN - Query Generation Checklist

Before searching, complete this checklist:

## QUEST Query Generation for "[TOPIC]"

### Q - Questions (5W1H)
- [ ] Who queries generated (2+ queries)
- [ ] What queries generated (2+ queries)
- [ ] Why/How queries generated (2+ queries)

### U - Universes (Stakeholders)
- [ ] User perspective query
- [ ] Provider/expert perspective query
- [ ] Critic/skeptic perspective query

### E - Expansions (Semantics)
- [ ] At least 2 synonym variations used
- [ ] Domain-specific terms included

### S - Scopes (Boundaries)
- [ ] Temporal scope addressed (current year included)
- [ ] Geographic scope if relevant

### T - Types (Sources)
- [ ] Academic/research query included
- [ ] Practical/industry query included
- [ ] User experience query included

**Total Queries:** [Count] (Target: 15-25)
**Coverage Score:** [X/12 checklist items]

Example: "AI Coding Assistants 2025"

QUEST Application

Q - Questions:

• Who: "AI coding assistant developers users", "GitHub Copilot creators"
• What: "AI coding assistant features capabilities", "code completion vs generation"
• When: "AI coding assistants 2025", "GitHub Copilot evolution history"
• Why: "AI coding assistant benefits productivity", "why developers use AI coding"
• How: "how AI code generation works LLM", "AI coding assistant integration IDE"

U - Universes:

• Users: "AI coding assistant developer experience review"
• Providers: "AI coding assistant enterprise deployment challenges"
• Regulators: "AI generated code licensing intellectual property"
• Critics: "AI coding assistant limitations accuracy problems"

E - Expansions:

• Synonyms: "AI copilot", "code generation AI", "AI pair programmer"
• Related: "autocomplete", "code suggestions", "intelligent code completion"

S - Scopes:

• Temporal: "AI coding 2024 vs 2025", "future of AI coding"
• Scale: "AI coding individual developer", "AI coding enterprise adoption"

T - Types:

• Academic: "AI code generation research arxiv"
• Industry: "AI coding assistant market report"
• User: "GitHub Copilot reddit review"

Total: 22 queries across all dimensions

Integration with Standard Mode

When to use: Phase 2: PLAN, after hypotheses are formed

Minimum requirements by tier:

Output: List of queries ready for parallel WebSearch execution in Phase 3: RETRIEVE