A FULL-STACK JOB PORTAL THAT MATCHES RESUMES WITH JOB DESCRIPTIONS TO PROVIDE PRECISE RECOMMENDATIONS

Abstract

Job recommender systems (JRS) are specialized information filtering tools designed to assist job seekers in identifying job opportunities that align with their skills and experiences. These systems help users navigate the overwhelming volume of job postings on platforms such as LinkedIn and Indeed. Despite various strategies implemented in JRS, many fail to recommend suitable positions when evaluating multiple job offers. A common limitation is treating skills as static entities tied to job descriptions, without dynamically aligning them with job seekers' profiles. This paper presents a content-based job recommender system aimed at addressing these challenges by analyzing resumes and job descriptions to suggest the top-n relevant jobs for candidates. The proposed system leverages content-based filtering to measure the similarity between the skills extracted from resumes and the explicit features in job listings. Data was collected through web scraping of job descriptions from Indeed, focusing on major Saudi Arabian cities (Dammam, Jeddah, and Riyadh). Key skills in demand were analyzed, and recommendations were generated by matching these skills with those in candidate profiles. The effectiveness of the system was evaluated using decision support measures to quantify recommendation accuracy and error rates.