Role of AI and Data Science in Cyber Security: a joint workshop UNSW/Data61

ONLINE | A workshop for researchers to facilitate new collaboration ideas at the intersection of data science and machine learning in cybersecurity.

We invite you to register and participate at this workshop  which will cover some exciting and emerging ideas on the intersections of AI and Cyber Security.

About this event

AI techniques have been used for managing cybersecurity threats for several decades. However, recent years have seen a range of sophisticated techniques that are being used for both defence and offence. Recent developments in machine learning have expanded the attack surface with many new types of attacks not seen before. In addition, the ML algorithms themselves pose new threats as they can be manipulated by adversaries.

Data remains a fundamental component of any cybersecurity research. Recent advances in data science allow processing and ingestion of large volumes of data in real-time, facilitating the creation, development and deployment of transformative data-driven decision-making tools that help us to address future cybersecurity challenges.

This one-day workshop provides a networking opportunity for UNSW and Data61 researchers, to highlight the diverse expertise and active projects from both institutions in the relevant areas. Ultimately, we seek to facilitate new collaboration ideas in the form of PhD proposals and other research projects at the intersection of data science and machine learning applied to cybersecurity.

This workshop includes a complimentary lunch voucher to use on the day.

Join in the conversation now at our Microsoft Teams Channel.

Speakers

Prof. Rakesh Verma

Phishing Through the Data Science and Human Lenses: Progress and Gaps.

Abstract: Phishing is a persistent threat that has caused incalculable damage. Much attention has been devoted to the detection of phishing emails, websites and links. Less attention has been given to the human angle, which is also important, since attackers are always adapting and creating new attacks. The challenge for models is to not only adapt to new attacks but also remember the classical ones. The challenges for humans are many. After almost two decades of research on phishing, what have we learned and where should we focus our energies? In this talk, I will consider both aspects of phishing research: technological and human. I will give an overview of phishing detection and human studies research, report on some of our experiments and work to push the envelope and systematize research on phishing, and highlight opportunities for future research.

Bio: Rakesh Verma is a Professor of Computer Science at the University of Houston (UH) where he leads the Reasoning and Data Analytics for Security (ReDAS) group. He won the 2013 Lifetime Mentoring Award from UH for mentoring undergraduate researchers, and is Program Co-Chair for the 11th and 12th ACM Conference on Data and Applications Security and Privacy (CODASPY 2021 and 2022). He is the co-author of Cybersecurity Analytics. He was an ACM Distinguished Speaker for seven years, and is a Senior Fulbright Specialist. He has a B.Tech. degree (highest honors, Gold Medal) from IIT Varanasi, and MS and PhD degrees from the State University of New York at Stony Brook, where he was a Catacosinos Fellow.

Prof. Elisa Bertino

AI-Powered Network Security.

Abstract: Networks are today a critical infrastructure. Their resilience against attacks is thus critical. Protecting networks requires a comprehensive security life cycle and the deployment of different protection techniques. To make defenses more effective, recent solutions leverage AI techniques. In this talk, we first discuss relevant directions for AI-based protection techniques, according to a security life cycle. We then present Polisma, a framework to learn access control policies from data; such an approach is critical to enable zero-trust architecture (ZTA). Polisma is based on a pipeline of different techniques to learn attribute-based access control (ABAC) rules from logs of access control decisions and potential context information obtained from external sources (e.g., LDAP directories). ABAC is being widely adopted due to its flexibility in capturing authorizations in terms of the properties (attributes) of users and resources. However, specifying ABAC policies is a complex task due to the variety of such attributes. Polisma, combines data mining, statistical, and machine learning techniques to learn access control rules that can then be easily understood by end-users, auditors, systems administrators. We have experimentally evaluated Polisma using two datasets (real and synthetic). Experimental results show that Polisma is able to generate ABAC policies that accurately control access requests.

Bio: Prof. Bertino is currently a Samuel Conte Professor of Computer Science with Purdue University, West Lafayette, IN, USA, and serves as director of the Purdue Cyber Center and as research director of CERIAS. She leads multidisciplinary research in data security and privacy. She is a Fellow of the IEEE and a Fellow of the ACM. She received the 2002 IEEE Computer Society Technical Achievement Award and the 2005 IEEE Computer Society Tsutomu Kanai Award for pioneering and innovative research contributions to secure distributed systems. She is co-editor in chief of GeoInformatica and has served as EiC of IEEE Transactions on Dependable and Secure Computing, and editor of the Synthesis Lectures on Information Security, Privacy, and Trust.

Ian Yip

How the cybersecurity industry uses AI

Abstract: Unfortunately, AI has become a buzzword in the world of cybersecurity. Most people don't understand what it means. As a result, they end up using it more for marketing purposes. But there are actual effective applications of AI in cyber defence, when used properly. This session will take you through the perceptions and practical uses of AI in the real world of cybersecurity.

Bio: Ian is the Founder and CEO of Avertro, a venture-backed cybersecurity software company based out of Sydney, Australia. He has held a variety of leadership, advisory, strategy, sales, marketing, product management, and technical roles across Asia Pacific and Europe in some of the world's leading companies including McAfee, Ernst & Young, and IBM. He is also a graduate from the School of Computer Science and Engineering at UNSW.

 

Program

08:30 -- 09:00 Registration

09:00 -- 09:10 Opening/Welcome

09:10 -- 10:00 Keynote I (Prof. Rakesh Verma, Univ. of Houston)

10:00 -- 10:20 Morning tea

10:20 -- 11:20 Keynote II (Prof. Elisa Bertino, Purdue University)

11:20 -- 11:30 Break

11:30 -- 12:30 Paper Session I:

  • Explainability of GNNs for cyber - Yang Song (UNSW)
  • Security of AI -- Sharif Abuadbba (Data61)
  • Network security and AI - Sanjay Jha (UNSW)
  • Security of data science - Andrew Feutrill (Data61)
  • Ransomware and AI - Ejaz Ahmed (Data61)
  • ML/NLP for vulnerability assessment - Chandra Thapa (Data61)

12:30 -- 13:10 Lunch

13:10 – 13:40 Industry Talk (Ian Yip, CEO at Avertro, Sydney)

13:40 -- 14:40 Paper Session II:

  • Human centric and AI - Marthie Grobler (Data61)
  • Privacy of ML - Rahat Masood (UNSW)
  • API as regulatory tool - Rob Nicholls (UNSW)
  • Fake news and social graphs - Jiaojiao Jiang (UNSW)
  • Phishing Detection and AI – Arindam Pal (Data61)

14:40 -- 15:30 PhD poster sessions

15:30 – 15:45 Break

15:45 – 17:00 Breakout/Brainstorm sessions

17:00 – 17:10 Closing