Resume Matcher

A tool for finding the best matching resumes for a given Job Description

Setup:

1. Download the resume dataset:
   https://www.kaggle.com/datasets/snehaanbhawal/resume-dataset

2. Download the JD dataset:
   https://huggingface.co/datasets/jacob-hugging-face/job-descriptions/viewer/default/train?row=0

3. Install required libraries

   pip install -r requirements.txt
   

4. Set up your configuration details in config.py (use the given ones as reference)

   • job_description_csv should hold the path to the JD dataset CSV

   • resume_data_dir should hold the path to the data directory (archive/data/data/) of the resume dataset containing the resumes

   • extracted_resumes should specify the output path for storing the json file with extracted resume sections from the dataset

   • top5_result should specify the output path for the csv file storing top 5 matches for each JD

   • result_dict should specify the output path for the JSON file holding the complete matching results

   • mapping_dict should be the path to the mapping_dict.json file in the repo

   • resume_headers should be the path to the resume_header.json file in the repo

   • jd_range_left and jd_range_right specify the beginning and ending list indexes for the sublist of the JD list for which u wish to do matching

Execution

1. Run the extraction script to extract sections from all resumes

   python extract_resume_sections.py
   

2. Run the main script for matching resumes to JDs

   python main.py
   

3. The top5 matches for each JD will be found in the file top5.csv. For complete matching results check out result_dict.json