Matching Talent to Opportunities: Resume Parsing Using Arabic NLP and Machine Learning

A career break refers to a designated interval wherein an individual voluntarily abstains from work; however, despite the existence of several returnship programs, women have notable challenges while attempting to reenter the job market. Thus, to address the above issue, this paper suggests the development of a system that combines machine learning (ML) and Arabic Natural Language Processing (Arabic NLP). The proposed model follows a systematic approach: resumes undergo collection, cleaning facilitated by the “tnkeeh” library, and subsequent removal of stop words. The K-Nearest Neighbors algorithm is then applied for final classification. The system evaluates clients, linking them to suitable returnship programs and cultivating community support. In the feature extraction phase, resumes undergo cleaning, Term Frequency-Inverse Document Frequency (TF-IDF) vectorization, and job recommendation based on identified returnship related keywords. For classification, the KNN classifier is trained on preprocessed data, leveraging TF-IDF for relevance assessment and recommending returnship opportunities based on input resumes. This promising strategy successfully handles the problem in hand associated with the shortlisting of resumes by finding the top three job sectors based on the résumé and thereafter provides recommendations for the most suitable returnships for the user. Hence, the results show that the accuracy of this methodology is 96.89%.