SECLE solves the labour market mismatch problem with AI

SECLE, which funds adult education and skills services for continuous learning, needed up-to-date information on current skills needs to proactively promote the match between jobs and skills in the Finnish labour market. The match between labour market needs and skills is of great societal importance.   

Foresight helps to allocate funding correctly

“We want to better inform our customers of the emerging skills and talent needs in our society. In doing so, we can help prevent the mismatch we currently have in Finnish society,” says Kirsi Heinivirta, SECLE’s Director.

Over the past few years, SECLE has sought to use modern knowledge management solutions to help the labour market better find the talent it needs and to monitor the effectiveness of the funding it provides. Now, the use of AI is also playing a major role in multi-vendor collaboration.

SECLE wanted to use a big data analytics solution to gain more dynamic insights into future skills needs by leveraging job advertisement data, investment data, research publication data and large language model data. Background analysis includes an analysis of the skills currently produced by education and training. What kind of skills needs can be identified in the near future?

We used HeadAI’s big data mining tool, Graphmind, to combine the skills, education and investment data needed for the project. Solita modelled the results extracted with Graphmind in Lakehouse and visualised the results as Power BI reports. The reports are published in SECLE’s Skills Needs Compass online service.

“What is special about this is that such extensive data-driven integration of multiple datasets and AI analysis hasn’t been done at this scale in Finland or the world in general – it’s a very advanced solution,” Anu Passi-Raune, CBO, HeadAI says. 

Thousands of job titles on the labour market

For the first time, the skills and competencies extracted from job advertisements on the Finnish labour market have been made publicly available. The skills and competencies were categorised into 130 occupational fields. However, as there are more than 8,000 job titles in the Finnish labour market, the compilation of skills proved more challenging than expected and required a lot of human effort.

In the future, it will be possible to expand the knowledge by analysing competencies directly through the job titles on the labour market.

“We are living in a time of change with the reform of continuous learning. 

Until now, we have been talking about occupations. However, while job titles can remain the same for a long time, the skills required for those jobs can change. Particularly in sectors undergoing structural change, the skill sets within a job title can change radically,” says SECLE’s Meriluoto.

AI is a good servant

When the skills data from the labour market were compared with the skills data from degree programmes, it became clear that the language used by the labour market differed significantly from that used by vocational institutions and higher education institutions in Finland. As a result, it wasn’t possible to carry out machine analysis to the extent that had been planned.

“For example, the labour market talks about teaching skills and the educational sector about pedagogical skills. A human being understands that they mean the same thing, but a machine doesn’t. The challenges encountered during the project provide important insights for further development. The key lesson is that AI is a good servant, but it still needs a master. In this project, the refinement and classification of the skill terms extracted by AI were partly carried out manually,” says Anita Karppinen, Solita’s Senior Project Lead.

The project required a lot of experimentation and development, as the aim was to provide new customer groups with useful information on new skills needs.   

Data shows where to invest

The work has brought a new perspective to planning in the education sector and the basis on which new study modules are established and student intake determined. Everything must be based on labour market demand.

“The data will enable education and training providers to analyse training needs more independently from different perspectives: Which sectors have a shortage of workers? Which sectors are disappearing?” says Meriluoto.

SECLE can now provide training providers and coordinators with up-to-date information on current skills needs in the labour market. This will enable education and training providers to adapt their existing provision and develop it to meet labour market demand. In the future, SECLE will also be able to use skills data to determine the priorities for its funding and serve its customers in a much more comprehensive way.

Visualised data has been well-received

The new service has received good feedback from customers and international recognition. The number of users of the service has increased. Customers have shown interest in utilising and further developing the service.

“The feedback from customers has been very positive. For example, career practitioners have said that the Skills Needs Compass is like a candy store. There are treats that weren’t available before,” says Meriluoto.

Customers have been surprised by the quantity and quality of information provided by the Compass, as well as its ease of use.

“The feedback is that it gives a quick overview of current skills needs for different occupations, which is particularly useful when customers need to be served quickly. The visualised data has been very well received,” say Heinivirta and Meriluoto from SECLE. 

The OECD’s Trends in Government Innovation 2024

Fostering Human-Centric Public Services presents the Skills Needs Compass as one of the globally significant public sector innovations that promote user-centred planning and anticipation of future skills needs. Read more.