Privacy-Preserving AI and Graphs Energy Digital Twins

Your function

Do you want to contribute to the future of sustainable energy systems while addressing fundamental challenges in privacy, trust, and explainability?

A 4-year full-time, fully-funded PhD position is available at the Knowledge in Artificial Intelligence (KAI) group and the User-Centric Data Science (UCDS) group of the Vrije Universiteit Amsterdam, in close collaboration with TU Delft and societal partners including the Municipality of Alkmaar and Arnhem Electricity Campus, as part of the NWO-funded KIC project FEDERATE:
Fair Energy Data Environments for Renewable Autonomous Twin Empowerment
.

The energy transition increasingly relies on data-driven digital twins that integrate information from municipalities, energy providers, grid operators, and other stakeholders. While sharing data enables better planning, optimization, and decision-making, it also raises important challenges regarding privacy, data ownership, regulatory compliance, and trust.

As a PhD candidate, you will develop novel AI and semantic technologies for privacy‑preserving knowledge sharing in federated energy‑data environments. Your research will focus on combining knowledge graphs, ontologies, knowledge representation and reasoning, machine unlearning, and formal reasoning to enable the selective removal of sensitive information while preserving the usefulness of shared knowledge.

Your duties You will investigate how formal forgetting techniques, such as uniform interpolation and ontology forgetting, can be adapted to large‑scale knowledge graphs and digital twins. In particular, you will explore:

• Modular ontology engineering and sensitivity modelling for energy‑data ecosystems
• Formal forgetting and machine unlearning techniques for knowledge graphs and ontologies
• Temporal unlearning mechanisms for enforcing data‑retention and privacy policies
• Privacy‑preserving reasoning and federated knowledge graph infrastructures
• Formal verification techniques to certify that deleted knowledge cannot be reconstructed through logical inference
• Neuro-symbolic or machine unlearning approaches for privacy preserve
• Auditable and GDPR‑compliant AI systems for trustworthy data sharing

The project combines theoretical advances in Knowledge Representation and Reasoning with real‑world validation in energy‑transition use cases. You will collaborate closely with researchers from Vrije Universiteit Amsterdam and TU Delft, as well as stakeholders from municipalities and energy innovation ecosystems.

Towards the end of your PhD, you will evaluate your methods in realistic pilot environments involving municipal and energy‑campus data, contributing directly to trustworthy and responsible AI solutions for the energy transition.

Our PhD students are expected to publish their research at leading venues in Artificial Intelligence, Knowledge Representation, Semantic Technologies, and Machine Learning, including AAAI, IJCAI, ECAI, NeurIPS, ICML, ICLR, ACL, EMNLP, KR, ISWC, ESWC, The Web Conference (WWW), and other top international conferences and journals.

As a university, we strive to cultivate a welcoming and supportive community for researchers while championing equal opportunities for all. KAI ((Use the "Apply for this Job" box below).) and UCDS () are internationally recognized research groups with a strong track record in Artificial Intelligence, Knowledge Representation and Semantic Technologies. Human‑Centred AI, Semantic Web technologies, Knowledge Graphs, and responsible data‑driven systems.

We value diversity in all its forms and believe that different perspectives strengthen our research and teaching. We actively encourage applications from candidates with diverse backgrounds and experiences.

We are looking for a highly motivated PhD candidate with the following qualifications:

• A Master’s degree (or equivalent) in Computer Science, Artificial Intelligence, Data Science, Information Science, Mathematics, or a closely related field
• Strong interest in Artificial Intelligence, knowledge representation and reasoning, and semantic technologies
• Background in one or more of the following areas:
  • Knowledge Representation and Reasoning,…