Social network analysis enhances job recommender system - PageRank algorithm in the field of employment

Abstract

This study proposes an improved job recommender system (JRS) that integrates the PageRank (PR) algorithm with social network analysis (SNA) to enhance the accuracy and personalization of job recommendations. Traditional JRSs, which often rely on simple rules or collaborative filtering, fail to fully incorporate job seekers’ social network (SN) data, resulting in less personalized and comprehensive recommendations. In this approach, social network data from job seekers and public datasets are collected, cleaned, and standardized. A social network is constructed to represent the relationships between job seekers and jobs, and the PR algorithm is applied to rank nodes, assessing job seekers’ influence and connectivity within the network. Personalized job recommendation lists are then generated based on job seekers’ career background, skills, and PR ranking results. Evaluation metrics show that the precision of career recommendations is 0.85, recall is 0.75, and F-measure values outperform traditional algorithms. By incorporating the PR algorithm and SNA, this study addresses the limitations of conventional JRSs, providing a more accurate and personalized job recommendation method that better meets the needs of both job seekers and recruiters.