Modern supply chains are complex and involve many separate companies interacting with one another. Data generated as part of this process, e.g., production schedules, delivery routes, demand for a particular product, can be used to improve its various stages, and contribute to reducing resource waste, improving forecasting and planning by synchronizing upstream and downstream supply chains. The benefit of data sharing among supply chain actors can indeed help optimise the whole supply chain in a more holistic manner. However, datasets are unfortunately owned by mutually distrusting companies, and the sensitivity of the data makes the companies reluctant to be on broad with data sharing and collaboration. This represents a missed opportunity that keeps value locked within non-communicating silos. This project aims to bridge this gap by generating and applying appropriate statistical data anonymisation and obfuscation methods that remove sensitive information and retain data features that are useful for supply chain optimisation. The PhD candidate will become familiar with both the data privacy literature, such as k-anonymization and differential privacy, and the supply chain literature; research into the data flows that occur in a supply chain; identify the sensitive/useful data features within these flows; finally, propose and evaluate novel techniques that balance privacy with data utility in the identified context. The project will be jointly supervised by academics in the Business School and in the School of Computer Science.
Keywords: Data privacy, data security, supply chain and information sharing
Desired applicant profile:
- Quantitative researchers
- Knowledge on one programming language is desirable, not required, as training will be provided
- Undergraduate degree in statistics
Continue with Facebook
