National University of Science and Technology Politehnica Bucharest (ROMANIA)
About this paper:
Appears in: INTED2024 Proceedings
Publication year: 2024
Pages: 7208-7217
ISBN: 978-84-09-59215-9
ISSN: 2340-1079
doi: 10.21125/inted.2024.1898
Conference name: 18th International Technology, Education and Development Conference
Dates: 4-6 March, 2024
Location: Valencia, Spain
Today's students and recent graduates face the difficult task of selecting the ideal career path.. As they are just starting their careers, their options are determined by their educational backgrounds, innate abilities, and interpersonal skills - all of which are influenced by their social environments. Factors influencing one's career choice include alignment with interests, competences, and academic skills, alongside elements like job satisfaction, benefits, security, and workplace environment.

Education strongly influences students' career aspirations, as universities prioritize employability, emphasizing practical skills for the job market. During their undergraduate studies, students are usually required to follow a mentorship or internship program. This is intended to provide them with a firsthand understanding of the labor market and to prepare them for full-time employment.

Hence, due to the ever increasing number of applications made especially for job-hunting, such as Indeed, LinkedIn or Glassdoor, university-industry cooperation finds itself posed to find suitable careers for future career-seekers.

We aim to further help by recommending the individual’s career path from a pool of existing employment positions, stored inside an ontology of professions, extracted from the Romanian Occupation Classification (COR). To reach this objective, a research was conducted to determine which kind of recommender system to approach. Ontologies are collections of data and information that investigate how entities are connected through their shared membership in similar classes or categories. By combining information such as personal interests, education, and correlations between personality types and skills, we determined that a hybrid recommender system is a suitable choice.

We propose CareProfSys, a recommender system based on ontological inferences that can determine the students’ best career choices, using a recommendation algorithm based on the HermiT inference engine. By extracting the user’s information from their social media profiles, their Curriculum Vitae and their responses to a personality test, the information can be shown in a set of recommendations on the user interface. Once the suggestions are displayed, the user can choose to enter Virtual Reality mini-games in which they can experiment with their given recommendations, giving them an insight on how a specific occupation is like in real life. While VR cannot truly imitate a complex work environment, given the right amount of development it can nonetheless give an impression of it: we believe letting potential career starters try out their job outcome would positively influence the choice that follows it.

Through the involvement of 27 high school students during a summer elective course, an investigation was conducted to gather data on the aspects of career counseling. The resulting dataset includes information about the value of career counseling, occupation suggestions, and a nuanced analysis of personal characteristics related to career trajectories. A targeted case study that recommends the Chemical Engineer career path will be showcased to explain this research. The results not only improve our comprehension of the dynamics of career counseling, but they also highlight the significance of continuous learning both in university and in the workplace, and the increased need for cooperation between employers and academic institutions.
Career guidance, recommendation, employability, ontology, COR, virtual reality.