Purpose of Project

  • Manual screening of thousand resumes at a time for hiring potential candidates is indeed a tedious job and time-consuming process and the Human Resource Department of the Organization may have a biased decision to filter candidates from dozens of resumes.
  • Thus, a system needs to design and implement that will help the Human Resource Development of the Organization to hire potential candidates and gain a strong team to work in the Organization.
  • Automated CV analysis can provide effective personality prediction for job applicant shortlisting with unbiased decisions.
  • The proposed system can reduce time and efforts for both employee who has been hired and recruiter who is hiring the employee using artificial intelligence techniques.
  • The proposed system can be used in various candidate recruiting industries such as IT, Retail, Healthcare, Government, Doctor, etc.

 

Problem Solution

 

Figure : System Architecture

  • Input Dataset : Resume Data from Kaggle (14,806 resumes) in CSV file extension
  • Dataset Loading : Loading the CV data into the system with CSV file extension
  • Data Preprocessing : Filtering out the null records, empty field, noisy errors in the dataset to avoid data complexity
  • Feature Selection : Extracting significant features such as work experience, skills, education, certifications from each CV record using pattern matching and string matching techniques of AI
  • CV Scoring : Applying weighted average method on the extracted features based on recruiter’s references and expert’s knowledge
  • CV Shortlisting : Displaying potential CVs suitable for the specific job role
  • System Output : Desire Job Applicants

 

Technical Specification 

  • Tool Used : Pycharm (Version 2020.3.3)
  • Computer Programming : Python
  • Package : Python 3.9 (64 bit)
  • Operating System : Windows 7 and above
  • Inbuilt Python Libraries : numpy, re, pandas

 

Results 

  • Potential CVs suitable for specific job roles are stored in the system so that recruiter can refer the skilled CVs whenever required and may call the job applicant for an interview basis.
  • If the job role does not match the skills of job applicants, a message indicating no job applicant available is displayed for the recruiter to know about the job role status.
  • The input CV data taken in the project consists of three job roles namely Java Developer, Python Developer, PHP Developer.
  • Highly skilled employees according to recruiter preferences on work experience, skills, education, certifications are shortlisted automatically from a large pool of CV data.
  • As the project is of type P-class (Polynomial Time Bounded), we get results of shortlisted potential CVs from large data (here in the project, 14,806 resumes) instantly in real-time which is a grand success of the application.

 

Conclusion 

  • In pattern matching and string matching methods, the regular expression approach played an important role in the feature extraction process to extract important parameters from CV data such as work experience, skills, education background, certifications.
  • The weighted average method was used to score the CV performance based on job title specification.
  • Artificial Intelligence provides adaptive learning and performs real-time operations on input data.
  • It saves time and human efforts for both the hired employee and the recruiter who is recruiting the employee.
  • The proposed application offers a fair approach to applicant evaluation for a particular job opening.
  • Automated CV analysis can be used in a variety of industries, including IT, law, medicine, government, retail, and health.
  • This project work further can be extended to all file extensions as input CV data in the future.
Skip to content
%d bloggers like this: