Digital Skills labelling in ESCO

In line with the European Skills Agenda and Commission’s effort in tackling digital skills gap, in October 2022 ESCO introduced Digital Skills and Knowledge Concepts labelling, which distinguishes the digital skills and knowledge competences that are part of the ESCO classification. 

In ESCO the definition of Digital Skills is based on the The Digital Competence Framework for Citizens (DigComp), which is a framework developed to provide a common language to identify and describe the key areas of digital competence. DigComp uses the definition of digital competence adopted in the Council Recommendation on Key Competences for Life-long Learning: 

Digital competence involves the confident, critical and responsible use of, and engagement with, digital technologies for learning, at work, and for participation in society. It includes information and data literacy, communication and collaboration, media literacy, digital content creation (including programming), safety (including digital well-being and competences related to cybersecurity), intellectual property related questions, problem solving and critical thinking. (Council Recommendation on Key Competences for Lifelong Learning, 22 May 2018, ST 9009 2018 INIT).

Content

• 1. Methodology
• 2. Results & Use Cases
• 3. Accessing the data 
• 4. Additional documentation

1. Methodology

The labelling of digital skills and knowledge concepts is done combining machine learning (ML) algorithms with human labelling and validation, structured in the following five steps:  

• ESCO skills and knowledge concepts are manually labelled based on The Digital Competence Framework for Citizens (DigComp 2.2).
• Manual labelling of online job vacancies collected from the database of EURES.
• Development of a machine learning model to classify ESCO skills and knowledge concepts as digital or not.
• The selected model is used to calculate the digital probability of ESCO concepts.
• Comparing the list of concepts manually labelled as digital (first step) with the concepts automatically classified as digital (third step).

The combination of the two methods facilitated the identification of skills requiring further discussion and manual review. It also helped filling gaps due to human mistakes or imperfections in the training dataset.

2. Results & Use Cases 

As a result of this procedure, 1,201 ESCO skills and knowledge concepts have been classified as digital. This includes: 718 skills, 475 knowledge concepts and 7 transversal skills.  

The digital labels for ESCO skills and knowledge concepts can benefit several ESCO implementers:

• Public employment services: identify digital skills.
• Education providers: suggest the right course for lifelong learning.
• Policymakers: formulate policies for a digital-proof society.
• Researchers: investigate patterns around digitalisation.

3. Accessing the data

ESCO digital skills and knowledge concepts are available in the ESCO Download Section. To access the list, users are required to download the classification in the desired version (recommended the latest one), and the file "digitalSkillsCollection" will be part of the downloaded package.  

Finally, ESCO portal offers the possibility to add filters; therefore, the ESCO digital skills can be accessed by adding the Digital Label under the filter section.  

4. Additional documentation

For further insights into the development of Digital skills labeling within ESCO , please visit the following links:

• Technical Report Digital Skills and Knowledge Concepts: Labelling the ESCO classification.
• ESCO Data Science team blog post on Digital Skills and Knowledge Concepts.