arsath-eng/Tamil-spell-checker

Tamil Text Checker

A comprehensive tool for checking Tamil text spelling, grammar, and overall correctness using multiple AI models.

Features

• Multiple model support:
  • Rule-based checking
  • Deep Learning analysis (using Indic-BERT)
  • Statistical analysis
  • Google Gemma integration
• Real-time error detection
• Spelling and grammar correction
• Confidence scores for each model
• User-friendly Streamlit interface

Screenshots

Main Interface

Example Selection

Analysis Results

Installation

1. Clone the repository:

git clone https://github.com/arsath-eng/Tamil-spell-checker.git
cd Tamil-spell-checker

2. Install dependencies:

pip install -r requirements.txt

3. Set up environment variables:

cp .env.example .env
# Edit .env with your GROQ_API_KEY

Usage

1. Start the Streamlit application:

streamlit run main.py

2. Access the web interface at http://localhost:8501

3. Choose your input method:

   • Enter custom text
   • Use example text

4. Select checking options:

   • Spelling check
   • Grammar check

5. Click "Check Text" to analyze

Model Details

Rule-Based Model

• Uses predefined Tamil dictionary and grammar rules
• Fast and deterministic results
• Best for basic spelling and grammar checks

Deep Learning Model

• Powered by AI4Bharat's Indic-BERT
• Handles complex language patterns
• Suitable for nuanced grammar analysis

Statistical Model

• Uses TF-IDF and Naive Bayes classification
• Trained on correct Tamil text samples
• Good for identifying unusual patterns

Google Gemma Integration

• Leverages Groq's Gemma 2B model
• Provides detailed language insights
• Offers correction suggestions

Example Use Cases

1. Basic Spelling Check:

Input: நான் பள்ளிக்கு சல்கிறேன்
Output: Spelling error detected in "சல்கிறேன்" - Suggested: "செல்கிறேன்"

2. Grammar Check:

Input: நான் பள்ளிக்கு செல்கிறார்கள்
Output: Grammar error - Subject-verb agreement mismatch

Project Structure

tamil-text-checker/
├── main.py
├── requirements.txt
├── models/
│   ├── __init__.py
│   ├── rule_based_model.py
│   ├── deep_learning_model.py
│   ├── statistical_model.py
│   └── google_gemma_model.py
├── data/
│   └── tamil_words.txt
└── .env

Dependencies

Main dependencies include:

• streamlit >= 1.24.0
• indic-nlp-library >= 0.91
• transformers >= 4.30.2
• torch >= 2.2.0
• tensorflow >= 2.13.0
• scikit-learn >= 1.2.2
• Groq API client
• python-dotenv

Contributing

1. Fork the repository
2. Create your feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

Acknowledgments

• AI4Bharat for the Indic-BERT model
• Indic NLP Library contributors
• Groq for the Gemma model API
• Tamil language experts who helped validate the rule sets