“RESUME SELECTOR” Using Pyspark and Hadoop

Abstract

Resumes are commonly used for recruitment of employees by companies, but the selection process is not automated and the technology used to store and evaluate resumes is outdated. Hence, there is a need to develop a system that can accommodate huge amount of resumes received by the company and process them in real time. So, our proposed system uses the capabilities of Hadoop Framework to store Terabytes of data in a cluster to improve the efficiency of selection process. Further, PySpark is used to process the data parallelly in a distributed environment which generates the result in an efficient manner. The proposed algorithm works on keyword-based search (KBS) to filter out all the required skills from resumes. Further, the aggregate weightage for each resume is computed and checked against a confidence level to select the resumes. Due to distributed and parallel computation, our system performs in a more efficient and accurate manner than the traditional systems.