The project is simple Applicant Tracking System using a neural network without any high-level framework like TensorFlow or PyTorch. The system matches the keywords from Job description dictonaries to applicants' resume and predicts the job category with confidence score.
Clean and preprocess raw resume text. Convert resumes to TF-IDF vectors (max 784 features) Predict job category from resumes using a multi-layer neural network Supports top-3 category suggestions with confidence levels Includes training functionality on custom labeled resumes
- Clone the repository
git clone https://github.com/hvrdhn/ATS.git
cd ATS
- Install dependencies
pip install -r requirements.txt
- "ATS.py" : Contains Neural Network class
- "trainATS.py" : Data preprocessing, text extraction and training neural network.
- "testATS.py" : loads the trained model and predicts job category with confidence score.