Periodic Reporting for period 1 - SKILLAB (SKILLAB: Monitoring The Demand And Supply Of Skills In The European Labour Market)
Reporting period: 2024-01-01 to 2024-12-31
The SKILLAB project aims at (a) identifying skills shortages and gaps in the labor market for different regions, sectors,time periods (b) supporting all kinds of enterprises in developing their human resource strategy, including both training of current employees, and skill shaping of future employees, (c) supporting recruiting enterprises looking for specific skills, (d) supporting EU citizens in developing their skills in an informed way and eventually finding a suitable job and e) aiming to find good candidates for emerging roles, but also to equip existing employees with the necessary skills to retain their jobs, thus improving job retainability and reducing employee churn.The project will meet its goals by designing and implementing an intelligent platform for managing labor market skills and skill gaps. The system will monitor and mine internet resources and EU initiatives to acquire and process meaningful raw data about existing and sought skills, and entities offering and seeking jobs. State-of-the-art IT technologies will be used to empower the platform and achieve its goals, including advanced statistical and network and data analysis, machine learning and competency mining. On the basis of such analyses, recommendation tools for Human Resource Management, educational profile upskilling and digital policies will be made available to users. The system will offer its findings in an easy, humancentric with SSH principles, ready to use, and ethically correct form to both enterprises and individuals. Such objectives are highly relevant to the objectives of the European Union for upskilling, reskilling and to the promotion of open science, inclusion, diversity and equality. In particular, it is expected that the project will contribute to the multifaceted assessment of the labor market status through its data acquiring and analysis tools. Continuous monitoring will allow for detecting current and future trends and gaps in requested and processed skills. The platform will offer a knowledge base for recommendation tools to develop skill strategies both for enterprises and individuals. The primary objectives of the project are directed towards four distinct target groups (EU citizens, educational institutes, industrial stakeholders, policymakers), with each group reaping different benefits from the usage of the SKILLAB platform.
In the first year of the project, the focus was primarily to accumulate data from various sources, building the database of SKILLAB and the associated Tracker Services. In addition, the responsible partners set the foundations for the skill analytics, skill identification, skill mapping and skill projection methodologies that should advance the state of the art approaches.
Overall, the progress so far is satisfactory, with multiple objectives being completed. The SKILLAB Tracker is completed and is able to retrieve data from various sources, including job portals, policy document databases, user profile databases, article repositories and software repositories. A database has also been created that stores the retrieved information, while skill extraction routines are being executed with a specialized tool developed in our research labs.
Also, the SKILLAB Modeler is well underway, with multiple active research fronts expected to produce results and publications in the first six months of 2025. These fronts include skill biodiversity frameworks, skill archetypes, prioritization of skills across ESCO levels, detection of emerging skillsets, use of Explainable AI for important skill identification, network analysis for skill co-occurrence and discovery of knowledge units in programming code.
Finally, the SKILLAB platform is being developed and is in a good state, with multiple services being up and running.
Overall, the progress so far is satisfactory, with multiple objectives being completed. The SKILLAB Tracker is completed and is able to retrieve data from various sources, including job portals, policy document databases, user profile databases, article repositories and software repositories. A database has also been created that stores the retrieved information, while skill extraction routines are being executed with a specialized tool developed in our research labs.
Also, the SKILLAB Modeler is well underway, with multiple active research fronts expected to produce results and publications in the first six months of 2025. These fronts include skill biodiversity frameworks, skill archetypes, prioritization of skills across ESCO levels, detection of emerging skillsets, use of Explainable AI for important skill identification, network analysis for skill co-occurrence and discovery of knowledge units in programming code.
Finally, the SKILLAB platform is being developed and is in a good state, with multiple services being up and running.
Regarding the results, the following indicative outcomes have been achieved:
1) Novel tool for skill and occupation extraction, leveraging LLMs
2) Extension of ESCO taxonomy with association rules mining
3) Biodiversity inspired framework for skill profiling
4) Use of Explainable AI for Skill Demand Analysis
5) Skills Prioritization via hierarchical voting algorithm in conjunction with Compositional Data Analysis
6) Use of TURF analysis for optimal skillset detection
7) Analysis of leadership in SE and skills differentiation across leading and non leading positions
8) Bayesian Networks and classic network analysis for skills profiling in EU policies
9) Scientific Software Development and associated skills
10) SE hiring process automation and curricula design
11) Skills identification through artifact mining and detection of skills-at-risk
12) Curricula alignment with labour market needs via LLM-powered skill mining and classification/prediction
1) Novel tool for skill and occupation extraction, leveraging LLMs
2) Extension of ESCO taxonomy with association rules mining
3) Biodiversity inspired framework for skill profiling
4) Use of Explainable AI for Skill Demand Analysis
5) Skills Prioritization via hierarchical voting algorithm in conjunction with Compositional Data Analysis
6) Use of TURF analysis for optimal skillset detection
7) Analysis of leadership in SE and skills differentiation across leading and non leading positions
8) Bayesian Networks and classic network analysis for skills profiling in EU policies
9) Scientific Software Development and associated skills
10) SE hiring process automation and curricula design
11) Skills identification through artifact mining and detection of skills-at-risk
12) Curricula alignment with labour market needs via LLM-powered skill mining and classification/prediction