RNN

Resume Classification using RNN

This project implements an end-to-end Recurrent Neural Network (RNN) to classify resumes into different job categories based on their content.

Project Structure

resume-classifier/
├── data/
│   └── resume_dataset.csv
├── models/
│   ├── resume_model.pth
│   ├── model_config.pkl
│   ├── preprocessor.pkl
│   └── training_history.npy
├── src/
│   ├── __init__.py
│   ├── data_preprocessing.py
│   ├── model.py
│   ├── train.py
│   └── predict.py
├── templates/
│   └── index.html
├── app.py
├── main.py
└── requirements.txt

Installation

pip install -r requirements.txt

Usage

Training

python main.py --mode train \
    --data-dir data \
    --epochs 20 \
    --batch-size 32 \
    --model-type rnn \
    --use-attention \
    --bidirectional

Web Interface (Flask)

python app.py

Then open http://localhost:5000 in your browser.

The web interface provides a clean, modern UI for classifying resumes with visual feedback and confidence scores.

Dataset

The model expects a CSV file with at least two columns:

• Resume: The text content of the resume
• Category: The job category/label

Model Architecture

RNN Model (LSTM/GRU)

• Embedding Layer
• Bidirectional LSTM/GRU Layer(s) with optional attention
• Fully Connected Output Layer

CNN Model (Alternative)

• Embedding Layer
• Multiple Convolutional layers with different kernel sizes
• Max Pooling
• Fully Connected Output Layer

Features

• Text Preprocessing: Cleaning, tokenization, vocabulary building
• Attention Mechanism: Improved text representation
• Early Stopping: Prevent overfitting
• Learning Rate Scheduling: Adaptive learning rate
• Top-k Predictions: Get multiple predicted categories with probabilities
• Modern Web UI: Clean, responsive interface for easy predictions