May 2019 SCS Awards

This team, including Surya Nepal, Anton Uzunov, Marthie Grobler, Chadni Islam, Brigitte Biscotto and Siqi Ma, coordinated the second and highly successful CSIRO Data61 and DST Cyber Security Summer School. Surya and Anton secured five international keynote speakers, four of whom proceeded to explore further collaboration opportunities. Marthie presented her collaborative work with Indianna University, a direct outcome of an MoU signed by a keynote speaker from last year’s event. Chadni introduced two student lightening talk sessions that involved liaising with 26 PhD students from 12 universities, receiving much praise. Marthie and Brigitte managed the event website and Brigitte took the lead on event planning. Siqi was the master of ceremonies, doing a fantastic job.

Francesco chaired the organising committee for Convergence, a Post Doctoral conference. The event was fully subscribed and included interactive workshops, networking opportunities and a conference dinner. Over 85% of attendees rated the conference five out of five.

Brigitte supports numerous activities and people across SCS including but limited to: event organisation; coordination of visitors and summer scholars; travel coordination; DNFC reporting. Brigitte always remains positive, friendly and helpful.

This team, including Nick Van Beest, Brian Lam, Francesco Olivieri, Badiul Islam, Gabriela Ferraro and Silvano Colombo Tosatto, drove internal research collaboration on the use of natural language processing to help with automated translation of legal documents to Defeasible Deontic Logic (DDL). The aim was to create high quality annotated datasets to experiment and train NLP parsers/tools. In conjunction with the University of Auckland, the New Zealand Building Regulations were used to create datasets compatible with DDL. as were the New Zealand Food Standards. Both datasets were used to create a new dataset to train a semantic parser for legal language. These activities have been used as a guideline for the RegTech technology program.

Nick’s collaboration with the University of Groningen, University of Stuttgart and University of Tartu on business process generation (AI planning), compliance and process variability resulted in five publications and an invitation to write a chapter in the Legal Informatics Handbook. His work with the University of Melbourne on business process repair resulted in three publications, and his work with the University of Hasselt on efficient approaches for IT auditing resulted in an invitation to a PhD committee as an external advisor to guide further research.

Dongxi received this award in recognition of three significant outcomes: 1) Collaborative Research Project (CRP) with Monash University, which developed a series of mechanisms to provide data confidentiality for data collected by IoT devices and provide computation on encrypted domain. The results were published at the 25th ACM Conference on Computer and Communication Security (CCS); 2) Boeing Secure Modular Internet of Things (SMIT) project, which delivered a rapid prototyping platform for testing IoT security solutions as open source. Users can easily build a secure IoT communication network over RaspBerry Pi and open lab 802.14.5 radio; 3) Dongxi is also working to standardise his cryptographic solutions before the public introduction of quantum computing makes public key encryption invalid.

This team, including Seyit Camtepe, Raj Gaire, Adnene Guabtni, Tariq Abbasi, Surya Nepal and Cheryl George were responsible for desiging the Data Airlock platform as part of a proposal for the Australian Federal Police (AFP), who provided $1.1M in funding for the project. The AFP capture large volumes of illegal material such as Child Exploitation Material (CEM) and wanted to make these datasets available to the research community to improve image classification methods towards automated image analysis. This posed legal and health-related consequences, which the Data Airlock platform addressed by protecting researchers from the data (rather than the other way around), ensuring AFP had full custody of the data at all times, and permitting operation of the platform on a semi-trusted organisation such as a university. The Data Airlock platform was identified by Nature Biotechnology as having scientific value and innovation, and also attracted attention from a range of government agencies. The Data Airlock is currently being redesigned as a distributed platform where sensitive data from multiple organisations can be jointly used for analysis without releasing or merging datasets.

This was the second project with the company Laava ID. Team members An Binh, Qinghua and Ingo Weber scoped customer needs before developing three designs and implementing one. A publication on the generalised architecture and its core components using the Laava project as a use case was shortlisted for ‘Best Paper’ award at the the ICSA. Christopher and Suhrid performed process mapping and feature analysis of capabilities of different technological solutions such as business rules, process conformance checking, on-chain and off-chain.

This team, including Hassan Asghar, Jonathan Chan, SirineM’Rabet, Farhad Farokhi, Ming Ding, Thierry Rakotorivelo, Paul Tyler and Engineering and Technology, worked with the Australian Taxation Office (ATO) to create a synthetic version of its longitudinal individual tax record dataset. The aim was to make it more widely accessible to deepen understanding of public policy and the national tax system, but preserving the privacy of individuals while retaining as much information as possible. The team contributed to delivering Provable Privacy Guarantees with the additional constraints of high dimensionality and longitudinal data, a longstanding challenge to the research community. They then went on to design and deliver a data sharing platform for the Commonwealth Government as part of the Expanding Access to Sensitive Data (EaSD) program in collaboration with engineering teams.

Raj was the engineering lead for the Data Airlock team and was instrumental in the delivery of the first version consisting of primary air-gapped communication protocol and 3-tier web interfaces. During a Federal Government inter-departmental Deputy Secretary working group meeting, Raj was named as going beyond expectations to ensure successful delivery of the project.

Trustworthy Systems (TS) is one of the largest groups in Data61. With many externally-funded projects running to aggressive schedules, June has been instrumental in creating and maintaining a culture of achievement by aiming high and fostering close collaboration to meet goals, while also encouraging diversity of thought and inclusion. TS continues to be successful due to June’s insightful leadership and cultivation of deep technical skills.