This project is a Language Detection Model built in Python.
It loads and cleans a multilingual text dataset, trains a machine learning model, and allows users to test language predictions on custom input.

The model uses TF-IDF Vectorization and Logistic Regression to classify text into its respective language.
Additionally, the project includes data cleaning, model evaluation, and visualization of results.

🚀 Fast Language Detection
Detects the language of any given text within milliseconds using a trained ML model.

🧹 Automated Data Cleaning
Removes duplicate or over-represented text entries to improve model accuracy.

📊 Model Evaluation & Reporting
Generates accuracy score, classification report, and detailed confusion matrix.

🎨 Beautiful Visualizations
Creates heatmaps, histograms, and line plots to visualize model performance.

💾 Model Persistence
Saves the trained model as a .pkl file for easy re-use without retraining.

🖥 Interactive Testing
Accepts user input directly from the terminal to test predictions instantly.

🛠 Customizable Training Pipeline
Easily adjust vectorization method, features, and classifier for experimentation.

📁 Multi-Language Dataset Support
Handles large multilingual datasets with millions of entries efficiently.

• Dataset Analysis & Cleaning

  • Loads CSV dataset
  • Removes repeated text entries (>3 times)
  • Filters languages with very low frequency

• Model Training

  • TF-IDF character-level features
  • Logistic Regression classifier
  • Accuracy score and classification report
  • Saves trained model as .pkl

• Visualization

  • Confusion matrix heatmap
  • True positives line plot
  • True positives histogram

• Model Testing

  • Interactive user input
  • Predicts language instantly

Vectorizer: TfidfVectorizer(analyzer='char', max_features=2500)

Classifier: LogisticRegression(max_iter=2500, solver='saga')

Split Ratio: 70% train / 30% test

Evaluation Metrics: Accuracy, classification report, confusion matrix

Clone the repository

git clone https://github.com/AUX-441/Language-Detector-Model.git
cd Language-Detector-Model
python Test_Model.py
Enter Your Language here (type 'exist' to quit): Bonjour
Predicted Language : French
---