Cyber Security CRC Scholarship Opportunities

UNSW Canberra is looking for talented researchers to join our team in solving challenging problems in Machine Learning, Natural Language Processing, Deep Learning and Network Dynamics for Cyber Deception working with the Cyber Security CRC €“ see https://www.cybersecuritycrc.org.au.

Successful candidates would be awarded a Cyber Security CRC research scholarships or post doctoral fellowships. For further details see: 

Cyber Cooperative Research Centre Graduate Scholarship.pdf

Cyber Cooperative Research Centre Postdoc Fellowship.pdf

UNSW Canberra has a long standing relationship with Penten €“ a Cyber Security CRC Industry Participant, and we are looking for motivated individuals to work alongside our industry partners and academics in the exciting field of Machine Learning and Cyber Deception

Deception is an increasingly important tool in modern cyber security. A comprehensive cyber deception strategy, populated with realistic content and behaviours, can attract and expose attackers. Deceptive traps can be placed amongst real assets to entice  intruders or malicious insiders, who reveal their presence and intentions by interaction with the traps.

The success of such traps, however, depends on the extent to which they realistically mimic the attributes of actual components, devices and data. Their content, behaviour, metadata and interactions must all be similar enough to their real counterparts, without exposing protected or sensitive attributes, to be able to trap the adversary.

Historically, successful deceptions have been hand-crafted by domain specialists, at considerable costs in time and effort. Our project addresses the cost impediment to large scale deception adoption by developing technology to automate the generation of fake attributes and data. We harness Machine Learning to create realistic traps by learning from data on the protected assets. The traps include documents, databases, devices populated with user profiles and associated file systems, network traffic and other assets and processes.

The challenges of automation from real data, while managing the exposure of protected information, raises research questions in a number of areas in Machine Learning, Natural Language Processing, Deep Learning and Network Dynamics. Quantifying the characteristics of traps and empirical evaluation of trap efficacy and their impact on legitimate users are also components of the ongoing research activity.

Applicants must be Australian Residents.

Supervisor(s)

Nigel Phair, Director, UNSW Canberra Cyber n.phair@unsw.edu.au