This publication has been prepared by the interagency technical vocational education and training (TVET) group on skill mismatch in digitised labour markets, to support experts and policy-makers who wish to engage in discussion on the potential of web-based big data for skills policy.
It outlines how such data can be used to mitigate labour market challenges, reduce skills mismatches and strengthen the links between the labour market and education and training.
The focus is on overcoming conceptual and practical challenges and limitations, system development and using big data for skills policy in practice. Examples of big data initiatives from around the globe illustrate its potential and provide insight into how big data are already supporting policy-makers in shaping the futures of work and education.
- Policy-makers need faster and more detailed information on skills to monitor and respond to the challenges created by structural economic and societal megatrends and the Covid-19 pandemic.
- Providing information in (quasi) real-time, online labour market data have great potential to improve policy-makers’ understanding of trends in skills needs and supply.
- The strengths of web-based big data include timeliness and granularity compared to conventional approaches to skills analysis.
- While web-based big data have significant potential for skills policy, they tend to require more effort to prepare for analysis than data collected using conventional approaches. The unstructured information provided often suffers from statistical, selection and conceptual biases.
- In low-income countries, web-based big data analysis can provide useful insights that complement conventional skills analysis, but biases can be more challenging. Higher informal employment and a less-developed digital infrastructure means online recruitment covers only a small part of the job market, particularly urban, formal and white-collar jobs. This complicates analysis that aims to cover the wider labour market.
- Despite advancements in information and natural language processing (NLP) and cloud computing, setting up a stable and well-functioning system for gathering, processing and analysing big data remains challenging. Developing such a system is a complicated and resource- intensive endeavour, but one that can pay off in the long-run.
- Web-based big data cannot and should not replace other skills intelligence methods and sources. Exploiting the complementarities of big data and other sources of skills intelligence is key in generating statistically robust, detailed, and policy-relevant evidence.
- It is the combination of artificial and human intelligence that will be key for further developing big data’s role in shaping effective technical and vocational education and training (TVET) and skills policies in the coming years.
Chosen excerpts by Job Market Monitor. Read the whole story @ Perspectives on policy and practice | Cedefop