INVITED TALKS
Invited Speaker 1:
Ana Rosa Cavalli (Montimage/Institut Polytechnique de Paris/Telecom Sudparis)
Edgardo Montes de Oca (Montimage CEO)
Cybersecurity, Monitoring, Explainability and Resilience
Abstract:
We introduce a comprehensive study focused on prominent cyber resilience methods. Resilience is defined as a system's capacity to function despite attacks or breaches. Our aim is to highlight techniques that guarantee system security and performance, even in compromised conditions. While detecting intrusions and attacks is vital, there's a notable absence of methods addressing accountability and system robustness. By merging monitoring with explainability and resilience methods, we foster intrusion detection and create protective strategies. Our resilience model emphasizes self-repair and introspection, drawing inspiration from techniques like moving target defence. We provide real-world examples to demonstrate anomaly detection, explainability and system robustness, including the detection of autonomous vehicle communication anomalies, explainable AI techniques, and an electric vehicle charging scenario.
Biography:
Ana Rosa Cavalli has obtained her Doctorat d'Etat es Mathematics Science and Informatics, from the University of Paris VII, in 1984. She has been professor at TELECOM SudParis since 1990, and director of the Software-Networks department from 2005 to 2015. She is also a member of the research laboratory CNRS SAMOVAR. She is now emeritus professor at Institut Polytechnique Paris/Telecom SudParis and she is the Montimage research director. Her research interests are on testing methodologies for conformance and interoperability testing, active testing and monitoring techniques, analysis and validation of cybersecurity properties and their application to services and protocols.
Edgardo Montes de Oca graduated as engineer in 1985 from Paris XI Orsay (University of Paris-Saclay) both in electronics and computer science. He has worked as research engineer in the Alcatel Corporate Research centre in Marcoussis, France and in Ericsson’s Research centre in Massy, France. In 2004, he founded Montimage, and is currently its CEO. His main interest are in building critical systems that require the use of state-of-the-art fault-tolerance, testing and security techniques; the development of software solutions with strong performance and security requirements; designing and building tools for monitoring the security and performance of networks; and building portable and secure 5G solutions.
Invited Speaker 2:
Xuyun Zhang (Macquarie University, Australia)
Recent Advances and Challenges in Membership Inference Attacks on Machine Learning
Abstract:
Recently, data privacy and machine learning model security have received increasing attention from both academia and industry given the wide deployment of machine learning models in many real-world applications and the strict data privacy and cyber security regulations and laws issued by many governments. Machine learning (ML) models have been widely applied to various applications, but recent studies have shown that ML models are vulnerable to membership inference attacks (MIAs). MIAs aim to infer whether a data record was used to train a target model or not, and can directly lead to a severe privacy breach. MIAs have been shown to be effective on various ML models and many defence methods have been proposed accordingly to mitigate MIAs. In this talk, we would briefly discuss the recent advances of MIAs and provide the taxonomies for both attacks and defences to inspire the researchers who wish to follow this area. Then, we present our recent relevant work about source inference attack and membership inference via backdooring.
Biography:
Dr. Xuyun Zhang is currently working as ARC DECRA research fellow, a senior lecturer, and the Course Director of Cyber Security in School of Computing at Macquarie University (Sydney, Australia). Besides, he has the working experience at University of Auckland and NICTA (now Data61, CSIRO). He received his PhD degree in Computer and Information Science from University of Technology Sydney (UTS) in 2014, and his MEng and BSc degrees from Nanjing University. His research interests include AI privacy and security, data privacy and cyber security, anomaly detection and big data mining, cloud/edge computing and IoT, etc. He has so far published 200+ high-quality internationally refereed publications in the relevant prestigious journals and conferences such as ACM Computing Surveys, IEEE TC, IEEE TPDS, IEEE JSAC, IEEE TKDE, IEEE TSE, IEEE TSC, IEEE TII, IEEE TDSC, IEEE TBD, IEEE TCC, IEEE IoTJ, IEEE ICDE, IEEE ICDM, AAAI, IJCAI, SIGIR, ACM WSDM, ACM CIKM, etc. The publications have attracted 8000+ citations (Google Citations) with h-index 50. He has been listed as one of the Clarivate 2021 Highly Cited Researchers. He has led or participated in 15 research projects with the total funding over AUD $5million. He is actively involved in professional services by serving several conferences as a general or program co-chair and several journal as an associate or guest editor. Besides reviewing submissions for top-tier journals, he has been regularly invited to join the technical program committee for top conferences such as SIGKDD, SIGIR, NeurIPS, AAAI, IJCAI, ICLR, TheWebConf (WWW), ICDM, WSDM, SDM, CIKM, DASFFA, etc.
Invited Speaker 3:
Janick Edinger (Universität Hamburg, Germany)
A Middlware for Multimodal Mobile Device Interaction
Abstract:
Accessibility plays a central role in the use of mobile devices. Mobile applications and operating systems are designed for high-resolution, touch sensitive screens that are controlled with fine motor finger movements. For a large number of people, this type of interaction with mobile systems is not possible due to physical impairments. This presents a significant challenge for individuals, as their diverse constraints make it infeasible to devise a single universal alternative interface. For this reason, we take the approach of a middleware for multimodal interaction with mobile applications. Our developed middleware is engineered to seamlessly accommodate various input modalities while and to translate actions into commands for the respective application, thereby fostering a more inclusive and accessible mobile computing environment. In user studies, we evaluated its potential to enhance mobile device accessibility and improve the digital experience for users with diverse physical impairments.
Biography:
Janick Edinger is an assistant professor of distributed operating systems at the Universität Hamburg, Germany. Janick studied at the Universität Mannheim, Germany, where he also received his PhD. He also studied at Taiwan National University and the University of Alberta, Canada, and completed research stays at the University of British Columbia, Canada, Hong Kong Polytechnical University, and Georgia State University, Atlanta, USA. Janick has published his work in proceedings of international conferences such as MobiQuitous, PerCom, IPDPS, ICCCN, IUI, MSWiM, COMPSAC, and CHIIR, and received the PerCom 2021 Mark Weiser Best Paper Award. His research interests include computation offloading, edge computing, Internet of Things, and assistive technologies.
Invited Speaker 4:
Jaehoon Paul Jeong (Sungkyunkwan University, Korea)
CBSS: Cloud-Based Security System with Interface to Network Security Functions
Abstract:
This paper proposes a Cloud-Based Security System (CBSS) with Interface to Network Security Functions (I2NSF) as the framework and interfaces. It shows the feasibility of CBSS for flexible and efficient security services in cloud-based network environments such as 5G networks and Internet of Things (IoT) networks. The design and implementation of CBSS are explained along with information and data models of the I2NSF standard interfaces. The architecture of the I2NSF framework is augmented for Intent-Based Networking (IBN) for intelligent security services. Through experiment, it is shown that CBSS can handle various security attacks autonomously.
Biography:
Dr. Jaehoon (Paul) Jeong is an Associate Professor in the Department of Computer Science and Engineering at Sungkyunkwan University since September in 2012. He got a Ph.D degree from the Department of Computer Science and Engineering at the University of Minnesota - Twin Cities in December 2009. His Ph.D advisors were Prof. David H.C. Du and Prof. Tian He. He got an MS degree from School of Computer Science and Engineering at Seoul National University in 2001. His MS advisor was Prof. Yanghee Choi. He got a BS degree from Department of Information Engineering at Sungkyunkwan University in 1999. Dr. Jeong worked for Brocade Communications Systems as a software engineer from January in 2010 to July in 2012. Before his Ph.D. study, he worked for Electronics and Telecommunications Research Institute (ETRI) as a member of research staff from February in 2001 to August in 2004. He has participated in the Internet Standardization for IPv6 Host DNS Configuration, Interface to Network Security Functions (I2NSF), and IPv6 Wireless Access in Vehicular Environments (IPWAVE) in the Internet Engineering Task Force (IETF) since 2002. His research interests include Internet of Things (IoT), Cloud-Based Security Service Systems, Vehicular Networks, Navigation Systems, and Indoor Localization.
Website: http://iotlab.skku.edu/people-jaehoon-jeong.php
Invited Speaker 5:
Aku Visuri (University of Oulu, Finland)
Wellbeing Insights in a Data-Driven Future
Abstract:
Our research explores the intersection of wearable technology and user-generated data, and obtaining meaningful insights. While wearables have become ubiquitous in monitoring health and well-being, the actual utility of the data they collect for end-users remains limited. Our TypeAware case study delves into users’ challenges in interpreting and deriving actionable insights from their wearable data. The TypeAware application aims to enhance user understanding of digital well-being and sleep quality data. Results indicate that, despite user engage- ment, participants encountered difficulties generating actionable insights from their data. Leveraging the capabilities of large language models, the research demonstrates the potential for automating insight generation, thereby transforming raw data into meaningful, user-friendly insights. Ultimately, this work calls for a shift in wearable technology design, advocating for more user-centric approaches that empower individuals to unlock the full potential of their wearable data for improved well-being.
Biography:
Aku Visuri is a senior researcher at the Center for Ubiquitous Computing located at the University of Oulu, Finland. His background is in designing and implementing solutions for companies in the healthcare business. He received his MSc in Computer Science and Information Networks from the University of Oulu in 2016 and PhD in mobile sensing in 2019. During his PhD work, the focus was on ubiquitous computing and quantified-self (QS). Specifically understanding users of QS applications and technologies, such as wearable devices, and designing new methods to enable users more efficiency. More recent work is focused on digital health and applying intelligent methods to contextual sampling and sleep tracking. He is currently working on the Academy of Finland SleepVention project.