Towards Interpretable and Actionable Provenance-based Intrusion Reports

PhD Studentship

Towards Interpretable and Actionable Provenance-based Intrusion Reports

PhD Studentship

The project: Computer systems are vulnerable. Not a day goes by without news of another data leak or security breach. Computer systems are massive, complex, human-created systems — and they are inherently flawed — we don’t have the technology to build perfect systems. Therefore, we need to develop a mechanism to respond quickly and accurately to intrusions. Currently, there is much research focused on detecting intrusions, which is a good start, but once we detect an intrusion, the immediate question is, “What is the root cause of the intrusion? What kinds of information are involved in it? How do we fix it?” This is the problem we aim to address, through sophisticated visualisation of the system execution. Our goal is to transform intrusion detection systems and data into a visualisation that makes apparent the right action to take.

Information is only meaningful if it can be communicated effectively. While there is a growing security community exploring provenance-based intrusion detection. However, the impact on the industry has been minimal. On the one hand there is mounting pieces of evidence that the capture of causality relationships in provenance graphs improves greatly over standard audit log format; on the other hand, the average human is not efficient in interpreting large and complex graphs. The student work will be vital in moving this body of work outside of the research community, by providing means to communicate the results effectively. We identify Three main objectives: 1. To study ML techniques used in automated provenance-based forensic and intrusion detection with the goal to identify methods to build more interpretable models. 2. To design effective communication vehicles targeted at well-defined audiences through graphical or textual means. 3. To study graph summarisation techniques to extract human-relevant information.

The group: The successful candidate will join the University of Bristol Cyber Security Group (UBCSG). UBCSG is recognised jointly by the National Cyber Security Centre (NCSC) and the Engineering and Physical Sciences Research Council (EPSRC) as an Academic Centre of Excellence in Cyber Security Research, and hosts a Centres for Doctoral Training in Cybersecurity. The successful candidate will join a dynamic and growing research and student community. They will have opportunities to work and collaborate with international partners in academia and industry.

How to apply

Prior to application if you are interested, please email ( with your CV and academic transcripts.

The formal application process can then be discussed.

Candidate requirements

First class in Computer Science or a related subject.

Basic skills and knowledge in at least one of Systems, Security, Graph Analysis, HCI, ML


Competitive scholarship covers full UK PhD tuition fees and a tax-free stipend (£22,000 in 2018 / 2019). The HM Government funding is restricted to British nationals.


Informal enquiries, please email Dr Thomas Pasquier,

General enquiries, please email

Application deadline


Thomas Pasquier
Lecturer (Assistant Professor)

My research interests include Digital Provenance, Operating Systems, Distributed Systems, Data Protection and Privacy, Internet of Things and Intrusion Detection.