VDAB Future of Work Challenge

VDAB launched the Future of Work Challenge in April 2019. Participants gained insight into the major challenges of the Belgian labor market and were asked to come up with innovative proposals for possible solutions.

33 companies and startups participated, 4 winners were chosen, among them Arinti with the CompeTrend proposal.

Why did VDAB hold the Future of Work Challenge?

It is important for us to work with an open mind. The Flemish labor market will face new challenges in the next 10 years. We suspected that with all the data we have available, innovative solutions are possible that we hadn't thought of ourselves. Companies and startups that are usually busy with other topics probably have a refreshing perspective. The purpose of the challenge was to encourage them to work out ideas.

Michael De Blauwe, Projectmanager @ VDAB Analytics & Innovation Team

How did the challenge work?

Interested startups and companies were given information about the problems on the labor market, about the data available to VDAB and about what could be gained with the challenge. 33 ideas for projects were submitted, from which VDAB eventually selected 4 winners.

Labor market forecasts

The skills mismatch leads to a growing underutilized potential and lower economic growth. Long-term unemployment threatens job seekers without the required skills.

This is why VDAB has developed four strategies to meet these challenges: (1) skills upgrading, (2) retraining, (3) activation and (4) productivity increase.

With our project we will help VDAB reach their objectives in terms of skills upgrading and retraining.

Competrend investigates which competencies will be popular in the labor market of the future. Based on Competent 2.0 (the most modern competency standard) we predict the future evolution of the competencies based on deep trend analysis from both CV’s and vacancies. We will also measure the comparability between different competencies. In our project, we will provide VDAB with an answer to the question: “Which already present competencies of ou jobseeker can serve as a stepping stone to learn new, future-oriented competencies?”. As a result, VDAB customers will find suitable trainings more quickly to retrain and reskill.


Trend Analysis


First of all, we will do deep trend analysis of the required competencies in both vacancies and CV’s:

  • Clearing seasonal cycles and random deviations from long-term evolution,
  • Predict short- and long-term evolution of competences via time series analysis (VARIMA, RNN, …),
  • Clustering of trends in different classes (disappearing, lowering, stable, …) to detect the trends of the future

Next, we will map the evolution of the mismatch between the requested competencies in the vacancies and the CV’s as the ratio of the number of vacancies to the CV’s with a marked competence.

Skill similarity 

Next, we’ll map data-driven similarities between different competences through association rule mining.
We’ll weigh the number of CV’s that contain both competencies by the number of CV’s that includes at least one of both competences. In the example above, the following rule can be detected from this:

  • if competency Python (or R) in CV, then competencies R (or Python) in CV with 70% probability: Skill similarity (Python, R): 0.8

Finally, hierarchical clustering of competences is performed based on similarities between the competencies themselves to improve the Competent 2.0 taxonomy.

Project results 

As a result of our project, we’ll help VDAB achieve two of its four core strategies, namely:

  1. upgrading skills by quickly giving VDAB customers insight into achievable skills to learn,
  2. retraining VDAB customers quickly and efficiently.
  • orienting jobseekers faster to appropriate and feasible training,
  • making the Competent 2.0 competency standard more insightful by: facilitating early discovery of competencies that will disappear in the future, mapping the relationships between competences, so that a better taxonomy can be drawn up, whereby competencies are clustered.
  • creating a better taxonomy and insights into competencies that can offer strong support to the VDAB training platform.