The project takes a deep dive into Applicant Tracking Systems (ATS) which form an integral component of the recruitment aspect of Human Resources (HR). The project looks into literature review (what these systems do, common approaches, discuss possible implications), human-centered argument and my research question and the experiments used to analyze this question. The analysis of the findings of the experiments were tested against a survey to provide validation for the arguments made and finally, the limitations and potential directions for further research are also discussed.
The Research Question
What is the dynamic between the aspects of efficiency and fairness when Applicant Tracking Systems are utilized into the pipeline of recruitment? What are the consequences of using ATS in the recruitment process? – identifying flaws in the system, exploring the expectations weighed on the applicant for tailoring their resumes
In specific, this repo deals with one of the experiments conducted in the project which is to craft a resume for a specific field using keywords commonly used in job descriptions of the field and test it against a specific description to see how it performs. The code published here was used for making wordclouds for job descriptions and skills for the field of software engineering.
Methods
- Data: A dataset (available as archive(6).zip in repo) of job descriptions was used from Kaggle. It is data from Dice.com, a prominent US-based technology job board, scraped from the website and published on Kaggle
- The data was cleaned and jobs pertaining to a specific position in the technical industry: “Software Engineer” were filtered
- Word clouds of the job descriptions and skills section were generated and a resume was crafted using these words