Abstract:

Although globalization, internationalization and current technological advancement came with an increased amount of world-wide career choices for students or young adults, new difficulties related to their employment have occurred, such as the incapacity of choosing the most suitable job or career path as soon as possible. Usually, careers are determined by education, either formal or informal. When students enroll to a faculty, they do this with a certain career path in mind, but sometimes the decision is not sufficiently informed, and the exploration of possible occupations is deficient. In these cases, a career counselor can come into play and provide guidance for proper planning. Unfortunately, in Romania, the career counselling centers are under-staffed, as one councilor provides such programs for more than 3000 students and is under-financed. Recommender systems or smart platforms could address this shortcoming, by providing customized educational and career advice. Ochirbat & all (2018) built OCCREC, an occupation recommendation for adolescents on interest, profile, and behavior based on both content-based and collaborative filtering methods. Razak proposes a career recommendation by asking fifteen questions about skills and interests of students and use a fuzzy logic method to give three possible careers (Razak & al., 2014). Other recommendation systems use LinkedIn user data (profile, working history, achievements) (Heap & al., 2014; Bostandjiev & al., 2013). Wang & al. (2014) implements correlation mining, clustering analysis and association rules for providing occupation recommendations to vocational students. Knowledge modelling techniques, such as ontologies, are used for enhancing recommendations. Hybridization of ontology-based recommenders with other recommendation techniques can enhance their performance even further, but the main challenge is to develop a highly expressive ontology which properly reflects a certain knowledge domain.

When talking about occupations, the reference ontology is the ESCO one, reflecting the "European Skills, Competences, qualifications and Occupations" taxonomy. Based on that, we propose a similar ontology reflecting the Romanian labor landscape, further called COR ontology. Our ontology contains all nine major groups of occupations, each of them having other four levels of categories, represented as hierarchy of classes. Jobs “per se” are present as individuals in our ontology. To assure inferences and recommendations, educational skills and competences are indicated for each job/ job category. The current article introduces the COR ontology graph, as compared to the ESCO ontology, as well as its description and some scenarios of utilisation for making recommendations of jobs and necessary educational steps to obtain those jobs from group 3 – “Technicians and other specialists in the technical field”.

Keywords:

Career guidance, educational recommendation, ontology, ESCO, COR.