danglive / resumefilter

Extracting relevant information from resume using deep learning.

Resume-Filter

Extracting relevant information from resume using Deep Learning.

Getting Started

Requirements

Code successfully runs with:

OS- Ubuntu 18.04
python 3.6.5
numpy 1.13.3
pandas 0.23.0
keras 2.1.5
pdfminer

Usage

Data Preparation

Getting training data is most challenging part due to lack of publicaly available dataset of resume. Currently, to create training data you will have to manually label the lines of resume.

1.Put the resume in data/training_data folder. Currently only .pdf and .docx format supported.

2.Run utils/create_training_data.py. A popup will be created. Annote the lines of resume in the same.

Training

3.Run start_training.py.

4.Trained model will be saved in trained_model folder.

Ready to use!

5.Put the resume to parse in data/resume_to_parse folder. Only .pdf and .docx format supported.

6.Run predict.py