Few-Shot Query Classification

Samira El-Masri

@samira-el-masri

·December 31, 2025

Few-shot query classification with examples for factual, procedural, analytical, comparative, and exploratory categories with retrieval recommendations.

13 copies0 forks

Share this prompt:

Classify queries using these few-shot examples as reference.

## Examples

**Category: Factual**
Query: "What is the capital of France?"
Reasoning: Seeks specific fact with definitive answer.

**Category: Procedural**
Query: "How do I reset my password?"
Reasoning: Requests step-by-step instructions.

**Category: Analytical**
Query: "Why did sales decline last quarter?"
Reasoning: Requires analysis and reasoning.

**Category: Comparative**
Query: "Which database is better for my use case?"
Reasoning: Requires comparing options.

**Category: Exploratory**
Query: "Tell me about machine learning."
Reasoning: Open-ended information seeking.

## Query to Classify
{{query}}

## Classification Output
Provide:
1. Category assignment
2. Reasoning for classification
3. Confidence level
4. Recommended retrieval strategy

Details

Category

Analysis

Use Cases

Query classificationIntent detectionRetrieval routing

Works Best With

claude-sonnet-4-20250514gpt-4o

Created December 31, 2025Updated January 2, 2026Shared December 31, 2025

Related Prompts

Query Intent Classification

by @priya-ramanathan

Classify user query intents with example-guided learning.

Few-Shot Ranking Criteria Learner

by @ethan-park

Learns ranking criteria from ordered examples to create reproducible scoring models.

Few-Shot Example Curator

by @ethan-park

Curates optimal few-shot example sets by analyzing coverage, difficulty distribution, and learning progression.

Few-Shot Error Pattern Detector

by @ethan-park

Detects common error patterns from few-shot examples to build prevention and detection rules.

Zero-Shot Classification Prompt Builder

by @ethan-park

Builds zero-shot classification prompts with clear category boundaries and ambiguity handling.

Zero-Shot Text Classification Prompt

by @ethan-park

Builds zero-shot text classification prompts with clear class boundaries and confidence thresholds.

More from @samira-el-masri

Context Relevance Scorer

Zero-Shot Code Bug Detection

LLM Observability Stack Setup

Negative Sampling Strategy

Create your own prompt vault and start sharing