Resume Classification using Elite Bag-of-Words Approach

As technology is advancing day by day, new trends are booming up, like automation, where traditional libraries are being automated to digital libraries. Therefore, instead of manually screening the resumes of all candidates, algoiithms and models are being employed to screen resumes in job and career portals. This complete process of mapping resumes to their corresponding job profiles could be efficiently accomplished by making use of various machine learning and Natural Language Processing (NLP) tools. This article utilizes a recently introduced text vectorization technique called Elite bag-of-wo rds for the vectorization of resumes. To implement this method, words in each dass are ranked based on their occurring frequency, and then applied maximum entropy partitioning (MEP) to derive the top-ranked significant keywords in each class. These keywords, defined as the Elite keywords, were extracted from each class, and concatenated without redundancy, for predicting the resume type. This research study presents an expeiimental comparison of the proposed method with existing bag-of-words approaches. This paper implements four vectorization techniques and it isproved that the Elite bag-of-words approach outperforms the othermethods for resume classification.