Soft and hard skills identification: insights from IT job advertisements in the CIS region

Labor market transformations significantly affect the sphere of information technologies (IT) introducing new instruments, architectures, and frameworks. Employers operate with new knowledge domains which demand specific competencies from workers including combinations of both technical ("hard") and non-technical ("soft") skills. The educational system is now required to provide the alumni with up-to-date skill sets covering the latest labor market trends. However, there is a big concern about the self-adaptation of educational programs for meeting the companies' needs. Accordingly, frequent changes in job position requirements call for the tool for in-time categorization of vacancies and skills extraction. This study aims to show the demand for skills in the IT sphere in the Commonwealth of Independent States (CIS) region and discover the mapping between required skill sets and job occupations. The proposed methodology for skills identification uses natural language processing, hierarchical clustering, and association mining techniques. The results reveal explicit information about the combinations of "soft" and "hard" skills required for different professional groups. These findings provide valuable insights for supporting educational organizations, human resource (HR) specialists, and state labor authorities in the renewal of existing knowledge about skill sets for IT professionals. In addition, the provided methodology for labor market monitoring has a high potential to ensure effective matching of employees.