BIG DATA MEETS APPLIED AI-CONGRESS 2018

The field of Big Data has matured and now the focus moves from 'typical' Big Data subjects like storage and interoperability to application, like predictive analytics but also pitfalls like bias. Applied AI connects with this subjects and also adds some topics: exlainability, reasoning, autonomy and human interaction.

Keynote presentations

Keynote by Prof. dr. Natali Helberger
Natali Helberger, Professor of Information Law from the University of Amsterdam, is a member of the Steering Board of the VWData program. She will discuss responsible value creation with data, and how Fair, Accurate, Confidential and Transparent (FACT) data science supports that.

Keynote by Prof. dr. Mireille Hildebrandt, “The rise of autonomous systems”
Mireille Hildebrandt is professor of Smart Environments, Data Protection and the Rule of Law at Radboud University Nijmegen, and research Professor at the group for Law Science Technology and Society (LSTS) at Vrije Universiteit Brussels. She will discuss the rise of autonomous systems and some of the legal implications, notably in case of safety hazards (critical infrastructure) and violations of fundamental rights (privacy, due process, non-discrimination). This concerns issues of liability for tech developers, those who put these systems on the market and the users, but also the question of legal personhood for AI.

Keynote by Prof. dr. ing. Philipp Slussalek
Philipp Slusallek is professor for Computer Graphics at Saarland University and Scientific Director at the German Research Center for Artificial Intelligence (DFKI), leading the research lab Agents and Simulated Reality. He will discuss the challenges related to training and validation of AI for the safe application of AI in the real world.

Break-out sessions in application areas

Defence and National Security
Hosted by: Adelbert Bronkhorst and Christiaan van der Berg
In safeguarding the security of our society and in conducting effective military operations, the availability of actionable intelligence is crucial. Deriving such intelligence from ever increasing amounts of data from sensors and various (open) sources is, however, a daunting task. One not only has to deal with ‘big data’ (Volume), but also with its reliability (Veracity) and with the different forms (Variety) in which it appears, and one has to process everything in time (Velocity). Successfully managing these four ‘V’s’ presents tremendous challenges which can only be overcome by applying sophisticated AI and by optimizing collaboration between AI and human users.

Dr. Selmar Smit (TNO) will present together with Peter de Kock (Pandora Intelligence) a perspective on present AI-implementations and experiments in the domain of National Security. Dr. Stephan Raaijmakers (TNO) and a representative from the ministry of Defence will highlight research and applications within the Defence domain.

Personalized Health
Hosted by Marcel van Zandvoort and Serena Oggero
In our society people tend to suffer from unhealthy life styles or from conditions that require lifestyle related coping mechanisms. Lifestyle changes are difficult to accomplish as they require long-term dedication. A solution is to offer a digital lifestyle coach that provides real-time advice.

Recent developments in Artificial Intelligence can improve such digital lifestyle coaches in many fields. For example, AI based digital coaches can learn from or adapt to personalized health data and reason about their obtained knowledge. These digital coaches will offer an improved personalized experience. The current state of the art in digital lifestyle coaches offer a plain advice. This advice can deal with quite complex health data, but lacks an explanation or argument towards the user, while such an explanation could improve trust of the user in its digital coach. Also, an explanation can function as a persuasion mechanism by offering clear and logical reasons that make sense to the user.

This session is opened by a presentation from Dr. Suzan Wopereis (TNO) about the value of AI, and Explainable AI in particular, for personalized digital health. This will be followed by a workshop from a domain expert in digital lifestyle coaches, and a demonstration of a sophisticated lifestyle coach that is capable of explaining itself.

Self-driving vehicles
Hosted by Hala Elrofai and Jan-Pieter Paardekooper
Due to the complexity of driving in the real world, AI presents an indispensable technology for the development of autonomous driving functions. Many of the practical difficulties coming from real-world driving put significant challenges on the development and use of AI-enabled technology in this domain. Examples include the trade-off between performance and speed; and the trade-off between comfort and safety. Other challenges relate to the need to guarantee meaningful human control and to provide proper timely explanations to the various users involved – such as the driver, road authorities and other actors on the road, for acceptability and safety.

This session is opened by Prof. Marieke Martens (TNO, TU Twente), who will discuss the need of a context-aware, self-aware and user-aware AI as enabler for self-driving vehicles. Two more presentations by experts in the field will be followed by discussion session on the challenges of AI for self-driving vehicles.

ICT and Cyber
Hosted by Freek Bomhof
Big Data and AI are used in many domains, but of course the ICT domain itself is also in the picture. An eye-catching point of this is cyber security. There is a lot of data involved in detecting illegal activities on the internet and setting up protection against these activities. Analyzing these data-flows, which are enormous, increasingly relies on AI and ML models. The application of these AI and ML models leads to new legal, ethical and technological challenges.

In this session, three perspectives are presented. The Dutch Cyber Security Center (NCSC) presents the way in which they improve cybersecurity in the Netherlands. TNO presents the scientific insights in the area of ethics and the system of law that are related to Malicious AI. Next, TNO presents the technological and European perspective from the current research on Dark Web and Virtual Currencies, more specifically how you collect data without having to drag a whole dragnet through the internet.

Small Big Data Congress
Hosted by: Harrie van der Vlag and Hans van Bracht
Small Big Data Congress with pitches with pitches providing an overview of practical and innovative applications based on big data. The pitches will be selected on the themes FACT (Fairness, Accuracy, Confidentiality and Transparency), and FAIR (Findability, Accessibility, Interoperability and Reusability). The call for pitches is now open: contact Mike Wilmer via mike.wilmer@tno.nl to present your Big Data research, dilemma, solution, application or model at the Small Big Data Congress!

Wrap-up and network event
The conference will conclude with a wrap-up by congress chair followed by a networking event with several Applied AI and Big Data demonstrations and poster presentations.

Looking for more information or want to enroll? Please visit the TNO website.