Intellectual analysis of the impact of job applicants' competencies on the level of remuneration offered to them

Introduction. The search for associative rules is one of the machine learning methods that allows you to detect patterns in data. The first attempts to implement it, related to marketing and advertising, were expensive. But with the development of information technology (IT) and the availability of a wide variety of data from the most different sources, the search for associative rules has become popular, including for labor market analysis. Online statistics of labor exchanges allows you to quickly track the demand for skills and knowledge of jobseekers, identify characteristics that affect the proposed salary level. Purpose setting. In this article, the purpose of the study is to identify associative rules between the sets of «proposed salary level» and «competencies» using the example of Product Manager vacancies. Methodology and methods of the study. To implement the search for associative rules was used Apriori algorithm, the assessment of support, reliability and lift indicators was carried out using the Deductor Studio platform. The collection of online data on Product manager vacancies from the HeadHunter website is implemented using the Python programming language. The main research methods are monographic, abstract-logical, bibliographic, intellectual and statistical analyses. The sample size is 282 vacancies. Results. The results of the study are of scientific and practical value for the development of labor market research tools. 15 associative rules have been identified for vacancies with a salary level of up to 50thousand rubles, from 75 thousand rubles and above (offers from this amount (the lower limit) are considered in the ads). For the remaining unexplored range of remuneration, a low quality of associative rules was obtained. Conclusion. The competencies highlighted in the associative rules can be divided into soft, providing communication, and hard, related to skills in the field of IT, economics, in particular finance, and statistical analysis. The patterns obtained show which skills will provide a higher salary. Low values of support for associative rules for vacancies from 50 to 75 thousand rubles are associated, in our opinion, with regional differences in wages. Intelligent job analysis allows not only jobseekers and employers to better navigate the labor market, but also educational organizations to respond to needs by adjusting curricula.