# End-to-End Resume Parsing and Finding Candidates for a Job Description using BERT

**Authors:** Vedant Bhatia, Prateek Rawat, Ajit Kumar, Rajiv Ratn Shah  
**arXiv:** [1910.03089](https://arxiv.org/abs/1910.03089)  
**Date:** 2019-09-30  
**Summary page:** https://www.summarizepaper.com/en/paper/1910.03089v2/

## Key points

- Authors address the challenge faced by employers in sorting through job applications to find suitable candidates
- Development of a resume parser to extract comprehensive information from candidate resumes
- Utilization of BERT sentence pair classification to rank candidates based on compatibility with a given job description
- Dataset includes resumes in both LinkedIn format and general non-LinkedIn formats
- Achieved 100% accuracy rate in parsing LinkedIn resumes and established a baseline accuracy of 73% for determining candidate suitability based on resume content
- Research offers a promising approach to streamlining recruitment process by automating candidate evaluation and ranking based on qualifications and experience outlined in resumes

## Summary

, , , , 


In their paper titled "End-to-End Resume Parsing and Finding Candidates for a Job Description using BERT," authors Vedant Bhatia, Prateek Rawat, Ajit Kumar, and Rajiv Ratn Shah address the challenge faced by employers in sorting through the increasing number of job applications to find suitable candidates. The authors propose an innovative solution that involves two key stages. Firstly, they develop a resume parser that is capable of extracting comprehensive information from candidate resumes. This parser is then made accessible to the public through a user-friendly web application. Secondly, the authors utilize BERT sentence pair classification to rank candidates based on their compatibility with a given job description. To approximate the requirements of the job description, they analyze the past job experiences detailed in each candidate's resume. The dataset used in this study consists of resumes in both LinkedIn format and general non-LinkedIn formats. The authors report an impressive 100% accuracy rate in parsing LinkedIn resumes and establish a strong baseline accuracy of 73% for determining candidate suitability based on their resume content. Overall, this research offers a promising approach to streamlining the recruitment process by automating candidate evaluation and ranking based on their qualifications and experience as outlined in their resumes. The combination of advanced technologies like BERT with efficient resume parsing techniques holds great potential for revolutionizing how employers identify top talent among a large pool of applicants.