Publication

Abstract : Finding and organizing suitable candidates for a vacant job position can pose challenges, especially when there is a high volume of submissions. This can impede team growth as it becomes challenging to identify the most suitable individual in a timely manner. To address this issue, an automated system called "Resume Recommendation and Classification" can expedite the selection process and aid in decision-making, while also reducing the time-consuming task of fair screening and shortlisting. This research utilizes Natural Language Processing to collect resumes and extract the necessary information. Subsequently, the resumes are ranked using BERT and TF-IDF word embeddings to obtain similarity scores. A classification model is then developed to categorize resumes based on job designations. To categorize the resumes, various techniques were employed, including Multinomial Naive Bayes, Linear Support Vector Classifiers, and k-NN Classifiers. These approaches enable the classification of resumes based on specific criteria.