Tijl De Bie - FEAST

Fair, Effective, and Sustainable Talent Management using Conditional Network Embedding

Description of the PI

tijldebie.jpgTijl De Bie is a Full Professor at the University of Ghent, having previously been affiliated to the KU Leuven, the University of Southampton, UC Berkeley, UC Davis, and Bristol University. He is most actively interested in the methodological foundations of data science, in approaches for ensuring fairness, transparency, and privacy in Artificial Intelligence, as well as in applications in music informatics, web and social media mining, computational biology, and human resources management. He has widely published in these areas, and his research was funded by several high profile research grants, most notably an ERC Consolidator Grant titled “Formalizing Subjective Interestingness in Exploratory Data Mining” (FORSIED), an ERC Proof of Concept grant titled “Fair, Effective, and Sustainable Talent Management using Conditional Network Embedding” (FEAST), as well as an FWO Odysseus grant titled "Exploring Data: Theoretical Foundations and Applications to Web, multimedia, and Omics Data".

Description of the project

The ongoing industrial revolution is the driver of a rapidly advancing shift in the division of labour between humans on the one hand and machines and algorithms on the other. This is the cause of significant challenges in the job market, such as the emergence of important skills gaps that need to be addressed by extensive upskilling or reschooling of workers. This requires considerable forethought and hence insight into the future job markets, as well as an understanding of how to best meet the job market's current and future needs. Substantial value is to be gained at all levels: from individual workers, over talent and human resources management within companies, to the determination of policy at governmental level.

This Proof of Concept proposal will address these challenges by leveraging results from the ERC Consolidator Grant FORSIED which lend themselves well to a uniquely suited and elegant data-driven approach. In particular, the method Conditional Network Embedding (CNE) offers a powerful framework for making sense of the diverse information relevant to human talent and the job market. It provides a platform to tackle a diverse range of use cases in a uniform manner. Moreover, a distinguishing advantage of CNE is that it offers mechanisms for compensating for existing biases in the job market, ensuring fairness, non-discrimination, and inclusion when deployed to these use cases.

During the project, a prototype platform will be developed, in tight collaboration with actors in the private and public sectors. This prototype will be evaluated, the IPR position investigated, and a market study conducted leading to a road to market strategy.