Skip to content

PROMPT_OPTIMIZER_README.md

Latest commit

 

History

History
185 lines (140 loc) · 5.58 KB

PROMPT_OPTIMIZER_README.md

File metadata and controls

185 lines (140 loc) · 5.58 KB
Raw

Enhanced Rule-Based Prompt Optimizer

A sophisticated React application that implements advanced rule-based techniques for optimizing AI prompts without requiring external APIs. This tool uses natural language processing to intelligently compress and improve prompts while maintaining their core meaning.

🚀 Features

1. Polite/Redundant Phrasing Removal

  • Identifies and removes verbose phrases like "Could you please provide me with..."
  • Converts indirect requests to direct instructions
  • Eliminates filler words and politeness indicators

2. Stop-Words and POS Tagging

  • Uses Compromise.js for natural language processing
  • Removes low-information words based on strategy (conservative/balanced/aggressive)
  • Identifies and removes adverbs, intensifiers, and filler words
  • Strategy-based word filtering

3. Passive Voice Conversion

  • Detects passive voice constructions
  • Converts to active voice for clarity and conciseness
  • Examples: "It is requested that..." → "Do..."

4. Expression Shortening

  • Replaces long phrases with shorter equivalents
  • Uses synonym mapping for common verbose expressions
  • Examples: "in order to" → "to", "provide a comprehensive explanation" → "explain thoroughly"

5. Structured Prompt Formatting

  • Converts multiple requests into structured formats
  • Uses bullet points and semicolons for clarity
  • Transforms questions into direct instructions

6. Keyword Extraction (Aggressive Mode)

  • Extracts essential nouns, verbs, and proper nouns
  • Filters out common words and stop-words
  • Creates minimal keyword-based prompts
  • Maintains action context (Explain, Summarize, Analyze, etc.)

🛠️ Technical Implementation

Dependencies

  • React 18 - Modern React with hooks
  • Compromise.js - Lightweight NLP library for browser
  • CSS3 - Modern styling with responsive design

Core Functions

// Main optimization function
optimizePrompt(prompt, strategy)

// Individual optimization techniques
removePolitePhrasing(text, optimizations)
removeStopWords(text, strategy, optimizations)
convertPassiveToActive(text, optimizations)
shortenExpressions(text, optimizations)
applyStructuredFormat(text, optimizations)
extractKeywords(text, optimizations)

Optimization Strategies

  1. Conservative: Minimal changes, removes only basic politeness
  2. Balanced: Moderate optimization, removes common filler words
  3. Aggressive: Maximum compression, includes keyword extraction

📊 Metrics and Analytics

The optimizer provides comprehensive metrics:

  • Token Count: Before/after token estimation
  • Quality Score: Algorithmic quality assessment (0.7-0.99)
  • Carbon Savings: Environmental impact calculation
  • Optimization Log: Detailed list of applied changes

🎯 Usage Examples

Input:

"Could you please provide me with a comprehensive summary of the economic impacts of climate change on developing countries along with key themes and main factors?"

Conservative Output:

"Provide summary of economic impacts of climate change on developing countries along with key themes and main factors"

Balanced Output:

"Summarize economic impacts of climate change on developing countries; key themes; main factors"

Aggressive Output:

"Summarize: climate change; economic impacts; developing countries; key themes; main factors"

🚀 Getting Started

  1. Install dependencies:

    npm install

  2. Start development server:

    npm start

  3. Open browser: Navigate to http://localhost:3000

📁 Project Structure

src/
├── components/
│   ├── PromptOptimizer.js      # Main React component
│   └── PromptOptimizer.css     # Component styles
├── utils/
│   ├── promptOptimizer.js      # Core optimization logic
│   └── testOptimizer.js        # Test cases and examples
├── App.js                      # Main app component
└── App.css                     # Global styles

🔧 Customization

Adding New Patterns

Extend the pattern libraries in promptOptimizer.js:

const politePatterns = [
  {
    pattern: /your-regex-pattern/gi,
    replacement: 'replacement',
    description: 'Description of optimization'
  }
];

Custom Stop Words

Modify the stopWords object to add domain-specific terms:

const stopWords = {
  conservative: ['your', 'custom', 'words'],
  balanced: ['more', 'words'],
  aggressive: ['even', 'more', 'words']
};

🧪 Testing

Run the test suite to see optimization examples:

import { testOptimizer } from './src/utils/testOptimizer.js';
// Run in browser console or Node.js

🌟 Key Benefits

  1. No External APIs: Completely offline and private
  2. Fast Processing: Rule-based approach is extremely fast
  3. Deterministic: Same input always produces same output
  4. Customizable: Easy to modify patterns and strategies
  5. Comprehensive: Multiple optimization techniques in one tool
  6. Metrics: Detailed analytics and quality scoring

🔮 Future Enhancements

  • Machine learning-based pattern recognition
  • Domain-specific optimization rules
  • Advanced grammar analysis
  • Multi-language support
  • API integration for cloud processing

📝 License

This project is open source and available under the MIT License.

Built with ❤️ using React and Compromise.js