ASF Winter School 2016

Big Data Anonymisation: An overview of existing models and solutions

While the benefits provided by new technologies are indisputable, they unfortunately pose considerable privacy threats that can potentially turn them into a nightmare. Most of these systems leave digital traces, or meta-data, that can potentially be used to profile/monitor users. Users are unaware of the information that is collected about themselves. Content on the Internet (documents, emails, chats, images, videos etc) is often disseminated and replicated on different peers or servers. As a result, users lose the control of their data as soon as they release them.

This talk will be an introductory course on Data Privacy, with a particular focus on Data Surveillance, Data Transparency and Data Anonymisation.

Speaker's bio: Claude Castelluccia is a research director (DR) at Inria Rhone-Alpes, where he leads the PRIVATICS team(Privacy Models, Architectures and Tools for the Information Society). The goals of the Privatics team are to study the new privacy threats introduced by the information society and design privacy-preserving solutions to prevent or at least mitigate them. The project follows a multidisciplinary approach. It focuses on technical and scientific problems, but also considers the economical, legal and social aspects of privacy. Claude has held visiting research positions at UC Irvine and Stanford University, USA. His current research is on Internet privacy and security with a focus on anonymized analytics, data anonymization, data transparency and Internet-based surveillance analysis. He is also interested in the economical and legal aspects of data privacy.

The Internet of Things is announced as the next big technological revolution where billions of devices will interconnect using Internet technologies and let users interact with the physical world. One of the key enablers for the IoT are wireless communications, to allow seamless interaction with billions of tiny autonomous integrated systems disseminated all around. Wireless Sensor Networks are crucial for turning the vision of IoT into a reality, and one major challenge is to provide multi-year lifetime while powered on batteries or using harvested energy.

In this talk, we will introduce the main challenges behind Wireless Sensor Networks. We will show how limited energy supplies have an impact on the whole stack, from hardware up to applications, and illustrate the many research challenges that stem from these constraints with work conducted in our group.

Reproducibility of Experiments in Concurrent Programming

Concurrent programming is the art of dividing a program into subroutines that threads can execute, hence exploiting the multiplicity of cores that today's computers offer. As in other scientific disciplines, reproducibility of experiments is key to the quality of concurrent programming research, however, most experiments involving concurrent programs remain non reproducible.

In this talk, I will present the challenges underlying the reproducibility of these experiments. Building upon the recent efforts from scientific venues to validate authors' accompanying material, I will present bits and pieces that should be disclosed to facilitate reproducibility. I will also present techniques to simplify this disclosure and finish with a hands-on session on Synchrobench, an open-source benchmark suite for concurrent programs.

Speaker's bio: Vincent Gramoli is a senior researcher at NICTA and an academic at the University of Sydney. Prior to this, Vincent worked on distributed systems while visiting University of Connecticut, MIT and Cornell University in the US. He participated in the development of the transactional memory stack at University of Neuchatel and EPFL in Switzerland. He worked on large-scale systems at INRIA, and obtained his PhD from University of Rennes and his Habilitation from UPMC Sorbonne University in France. His current research interest is on theoretical and practical aspects of distributed computing.