Over the last century, different learning theories have shaped the world of education and training before shifting to the competency-based approach (CBA). This new paradigm to teaching and learning aims to ensure that every student has to graduate with the competitive competencies of lifelong learners and is ready to enter the workforce and begin functioning in entry-level positions. However, despite the growing interest in competency-based learning and training, this field still faces numerous challenges, essentially the lack of consensus about an interoperable description of competency evidences. Indeed, the move towards CBA has created a need for effective instruments that support and assess competency development. In this context, the electronic portfolio (e-Portfolio) emerged as a suitable tool that helps learners collect and manage multiple kinds of assessment evidences linked to the program’s competencies from multiple sources. In this paper, we propose an ontology-based approach to e-portfolio modeling which relies on Semantic Web technologies to formally and semantically describe portfolio artifacts that evidence the achievement of one or several competencies. The proposed ontology is structured according to official e-Portfolio specifications, namely IMS ePortfolio and JISC Leap2A. In addition, other existing approaches to e-Portfolio modeling reported in the literature have been explored to avoid misinterpretation of these specifications. Furthermore, a comparative study of common e-Portfolio systems has been carried out in order to gain a fairly accurate idea of the generic structure of an e-Portfolio.