Automatic Software Engineering Position Resume Screening using Natural Language Processing, Word Matching, Character Positioning, and Regex
Dipendra Pant, Dhiraj Pokhrel, Prakash Poudyal
Abstract
Screening candidates' resumes manually is a tedious job, with possibilities of sometimes missing good candidates due to human errors, nepotism, and bias. However, these kinds of mismanagement don’t apply to machines. Instead, automatic screening of candidates reduces a lot of effort, time, and cost. Hence this work specifically focuses on extracting technical skills using natural language processing specifically resume label character positioning, data set consisting of software engineering candidate requirements, regular expressions, and word and phrase matching for candidate information retrieval. Character positioning a new technique for information extraction is introduced, which perceives needed data and pulls it out. This methodology creates a summary of the resume from the extracted information. And computes count scores from based recognized skills, and education plus experience level. Finally, upon testing on five random software engineering positions resume correct extraction rate of 33.59% was obtained.