Challenge
This project is part of the Fairness Innovation Challenge delivered by the UK Government’s Department for Science, Innovation and Technology (DSIT), and Innovate UK, in partnership with the The Equality & Human Rights Commission (EHRC) and The Information Commissioner’s Office (ICO).
The main motivation for our project stems from the need to address bias and discrimination in Artificial Intelligence (AI) systems, specifically in the context ofHigher Education (HE). Universities in the UK are increasingly turning to AI applications, especially Predictive Learning Analytics, to meet regulatory targets set by bodies like the Office for Students. The project is motivated by the aspiration to ensure that AI technologies, particularly those supporting student journeys, are fair and inclusive, irrespective of students’ diverse backgrounds or characteristics.
The market opportunity behind our solution is significant, especially considering the widespread adoption of AI in HE institutions. While many universities are beginning to explore AI applications in learning analytics, the sector lacks comprehensive strategies for incorporating fairness into these systems. Our innovation capitalises on this gap, providing a socio-technical approach to fairness in Predictive Learning Analytics, thereby presenting a unique market proposition.The market opportunity extends to institutions looking to enhance student success, improve retention rates, and meet regulatory compliance while fostering inclusivity.
In our research, we have identified similar innovations in the field, but they oftenlack a comprehensive socio-technical approach and may not address the uniquechallenges of HE environments. Existing solutions may focus solely on technicalaspects, overlooking the broader societal context in which AI systems operate.
Our project builds upon the foundation laid by the Open University’s existing Predictive Learning Analytics system, known as OU Analyse. This system, currently one of the largest globally, has been in operation for over a decade, offering valuable insights into the complexities and challenges of implementing such solutions. The work already done by the Open University in developing and deploying the existing system provides a strong basis for expanding its capabilities in response to the identified need.
