DPhil in Energy Data Privacy
Supervisors: Philipp Grunewald, Zeynep Tekler, Tina Fawcett
The Energy Demand Observatory and Laboratory (EDOL) is collecting personal and sensitive data relating to household energy use. At present it is difficult to share these data. High resolution data can contain personal insights that neither the data subject, nor researchers are aware of at the time of obtaining ‘informed consent’. Current best practice is to aggregate at least ten households. Why 10? Nobody knows. Ten may be too many - most research relevant information about diversity of demand may be lost. At the same time it may be too few. If nine of the homes are unoccupied, personal information about the tenth might still be disclosed.
This DPhil offers an exciting opportunity to develop a new standard for energy data processing and sharing. First, the exact nature of privacy concerns needs to be better understood. Open questions are:
- What type of information are data subjects concerned about disclosing, and are there other features that they are not concerned about, but should be?
- How can data be processed, such that concerns in (1) are addressed?
With the data available to EDOL it will be possible to train sophisticated models to generate synthetic data. These could potentially overcome privacy concerns while maintaining the relevant detail for different stakeholders.
If you are interested, please fill in this short form below, or apply formally here.
