In order to obtain data potentially employable in the DISKOW project, IER SAS approached a private company – Profesia, administrating the dominant online job vacancy web in Slovakia. In exchange for extraction from their database of online job vacancies, the IER SAS agreed to explore the modelling options of predicting the attractiveness of online job vacancies published by Profesia. This task was of particular interest for Profesia, exploring the analytical background to further improve their services. The cooperation of IER SAS with Profesia not only provided the DISKOW project with the necessary data but also allowed the Slovak partner to delve deeper into the topic and test analytical skills needed for the last part of the project – evaluation of outputs.
On September 23rd, researchers from IER SAS presented the results to representatives of Profesia. In the first part of the online meeting, presentation explaining the setup, methodology and results was delivered. Multiple predicting variables were identified using various designs of predictive models, out of which Random Forest yielded the most favourable results, showing the lowest erroneous indicators. As for the strongest predictors, for example, information on offered salary, required education and length of the job title showed a significant impact on the attractiveness of the vacancy note. In the second part of the meeting fields of potential cooperation between IER SAS and Profesia were discussed, concluding with an agreement to stay in touch and open for future cooperation.