Skip to main content

project

A matter of wording: Building a transversal skills taxonomy

1 July 2021 - 30 September 2021

PI/s in Exeter: Dr Georgina (George) Tarling

Research partners: Co-investigator: Dr Ana Melro, Instituto Universitário da Maia, Portugal.
Research Partners: Dominic Murphy and Brian Conway (Geek Talent)

Funding awarded: £ 5,000

Sponsor(s): ESCR-NPIF-Accelerating Business Collaboration

About the research

Taxonomies are widely used to predict emerging skills that will be needed in the rapidly changing jobs market. In the past such taxonomies were developed by panels of experts. Increasingly now they are created using machine learning techniques applied to large datasets. This has implications for education, as tools based on these taxonomies are increasingly used by policymakers and education providers to guide decision-making about vocational education and student employability. The processes used to identify and classify skills therefore warrant critical attention.

This project is a collaboration between the University of Exeter and Geek Talent, an analytics company in the North West, which uses natural language processing (a type of machine learning) to provide users (including students, job-seekers and education providers) with timely and detailed information to help them see which skills are needed in particular occupations, sectors and regions. The aim of the project is to help Geek Talent improve the quality of their Skills DNA tool, by comparing their data insights with the well-established and large scale ESCO (European Skills, Competences and Occupations) framework. From a business perspective, this will help give more detailed, meaningful and robust information to users. From an academic perspective it offers a way of critically interrogating two different platforms which both seek to shape education provision.

The outputs of the project include an updated taxonomy of transversal (including transversal digital) skills and a report describing in detail our methodology, a version of which can be uploaded to the Geek Talent platform to increase transparency and accountability. These outputs will provide a basis for funding applications to conduct ongoing research to improve the quality of machine learning generated taxonomies of emerging digital, data and other advanced technical skills. This will be of value particularly for the UK’s Institutes of Technology, who have responsibility for developing new FE and HE courses to meet the needs of regional employers with emerging technical roles.