The rapid growth of the data economy and the privacy implications accompanying it have motivated a new paradigm shift towards decentralisation of data on the Web, which aims to foster data sovereignty and empower individuals by allowing them to take back control over their data. However, the process of implementing data governance in a decentralised setting, especially when AI is involved, brings forth various socio-technical challenges related to data interoperability, process transparency, legal compliance (e.g. with GDPR, EU AI Act), privacy-preservation, and individuals’ trust and comprehension of decentralisation itself.
This project aims to address key questions related (but not limited) to:
Infrastructure: What data infrastructure is needed to enable the robust and efficient implementation of decentralised personal data governance frameworks?
Technology: How can multimodal knowledge graphs in combination with generative AI help individuals govern their data (e.g. within decentralised settings)?
User Empowerment: How to support individuals in making sense of their decentralised data and its governance? Can visualisations help?
Expected outcomes: A novel approach combining multimodal knowledge graphs, LLMs and visualisations to support individuals with their decentralised personal data governance. Raised awareness of data privacy and trust in AI and decentralisation. Scientific paper(s) publication at top tier ranked international conferences and journals in Computer Science.
Person specification:
- MSc in Computer Science or AI (minimum of 2:1 degree or higher)
- UK home student status holder
- Excellent oral and written communication skills in English
- Strong problem-solving abilities
- Interdisciplinary computer science interests
- Ability to work independently and in a collaborative environment
Essential skills:
- Experience with ontologies and knowledge graphs
- Experience with data privacy and interest in data governance
- Familiarity with, or keen interest in, decentralised technology (e.g. SOLID, distributed ledgers)
Desirable:
- Experience with generative AI (e.g. LLMs)
- Interest in Human-Computer Interaction
- Interest in privacy enhancing technologies (PETs)
Other:
- Experience in presenting or preparing scientific papers for journals and conferences is preferred but not essential.
Funding notes:
We are looking for an ambitious and motivated student with background in Computer Science or AI to undertake a 3 - 3.5 year fully funded PhD research project at the School of Computer Science, University of Birmingham. This scholarship covers (UK home) tuition fees and stipend. The stipend is the national UK standard.
We particularly encourage applications from students who have a background in ontologies, knowledge graphs and data privacy.
References:
- A. Kurteva and H. J. Pandit. “Relevant research questions for decentralised (personal) data governance”. In: Trusting Decentralised Knowledge Graphs and Web Data (TrusDeKW) Workshop at ESWC 2023, 28 May - 1 June 2023, Hersonissos, Greece.
- A. Kurteva and J. Domingue. “Towards Cultivating Decentralised Data Privacy, Interoperability and Trust with Semantic PETs and Visualisations”. In: NXDG: NeXt-Generation Data Governance, SEMANTiCs 2024, 17-19 Sep 2024, Amsterdam, The Netherlands.
- A. Meroño-Peñuela, E. Simperl, A. Kurteva, and I. Reklos. "KG. GOV: Knowledge graphs as the backbone of data governance in AI." Journal of Web Semantics 85 (2025): 100847.
This scholarship covers (UK home) tuition fees and stipend.
