As France's leading ecosystem in data science and artificial intelligence, the DATAIA Institute develops cutting-edge research by combining the skills of its 17 members of the Paris-Saclay cluster.

 

Two years after its launch, it is organizing a conference on data science, AI and society, to present the scientific advances made possible by its partnerships, as well as the prospects for industrial and international collaborations.

Several themes will be addressed: from innovations to the ethical challenges posed by AI, as well as the expectations of industrialists, or the various initiatives launched to animate the AI community of Paris-Saclay.

 

A poster session is also planned during the lunch-break. The call is open until 25 February 2020.

 
9h00 - 9h30
Welcoming
9h30 - 10h00
Introduction speech
10h00 - 11h00
Presentation of the DATAIA Institute
11h00 - 11h30
Coffee break
11h30 - 12h15
Keynote 
12h15 - 12h45
Presentation of two research projects
12h45 - 14h00
Lunch break
14h00 - 14h45

Round table « Ethics and AI »

14h45 - 15h30

Presentation of three research projects

15h30 - 16h00
Coffee break
16h00 - 16h45

Round table « Research, training and industry »,

16h45 - 17h30

Keynote: «Computational Fact-Checking:
A Selection of Problems and Solutions»

17h30

Closing cocktail

 

The DATAIA Institute invites students, doctoral students, post-doctoral students and researchers from the Paris-Saclay cluster to present their research work in the form of a poster.

The proposed work must address at least
one of the four interdisciplinary challenges of the Institute:  

 FROM DATA

TO KNOWLEDGE, FROM DATA

TO DECISIONS 

 MACHINE LEARNING AND ARTIFICIAL INTELLIGENCE 

 TRANSPARENCY, RESPONSIBLE AI

AND ETHICS  

 DATA PROTECTION, REGULATION,

AND ECONOMY 

Poster proposals must be submitted by February 25, 2020 in the following form:  

TITLE 

AUTHORS

ABSTRACT 

Please send your proposals to submission-dataia@inria.fr

A poster template is available below.

Posters should not exceed the A0 format (841x1189mm) and they will be printed by the authors.

 
 

Bertrand Braunschweig

Bertrand Braunschweig is director of Inria’s coordination mission of the national AI research programme.

President of the French Artificial Intelligence Association for four years, he joined the French National Research Agency in 2006 as head of several programs, then from January 2009 as head of the STIC department. He was also Director of the Inria Rennes - Bretagne Atlantique research center from 2011 to 2015, and Inria Saclay Ile-de-France from 2016 to 2018. In 2016, Bertrand Braunschweig coordinated the drafting of the Livre blanc sur l’intelligence artificielle, published by Inria.

Bertrand Thirion

Bertrand Thirion is head of the Parietal team (Inria Saclay - Île-de-France and CEA) and Director of the DATAIA Institute. After a thesis at Inria Sophia-Antipolis center under the direction of Olivier Faugeras, and a post-doc at the CEA Saclay, he participated in the creation of the Neurospin research center (CEA) where he created the Parietal team. His research focuses on statistical modeling and machine learning applied to brain imaging data. He is the author of more than 200 articles and contributes to the animation of open-source software projects in Python (scikit-learn, nilearn).

Executive board members 

Aapo Hyvarinen

Aapo Hyvarinen studied undergraduate mathematics at the universities of Helsinki (Finland), Vienna (Austria), and Paris (France), and obtained a Ph.D. degree in Information Science at the Helsinki University of Technology in 1997. In 2008, he was appointed Professor at the University of Helsinki. From 2016 to 2019, he was Professor of Machine Learning at the Gatsby Computational Neuroscience Unit, University College London, UK. Aapo Hyvarinen is the main author of the books "Independent Component Analysis" (2001) and "Natural Image Statistics" (2009), and author or coauthor of more than 200 scientific articles. Google Scholar gives him approximately 40,000 citations. His current work concentrates on unsupervised machine learning and its applications to neuroscience.

 

Bad Nudge-Bad Robot ?

Connected devices, specifically conversational agents such as Google Home or social robots, bring speech as a new dimension to interaction. Users tend to anthropomorphize these devices which could be soon able to detect user’s emotions and then will become a means of influencing individuals. They are currently neither regulated, nor evaluated and very opaque.

This project is studying nudges in human-machine verbal interaction, to understand the societal impact of these new devices. The ultimate objective of the project is to create «ethic-by-design» systems and also to create metrics of evaluation. This project is linked to the AI chair HUMAAINE.

 

Project directors: Laurence Devillers (LIMSI-CNRS, Université Paris-Sorbonne),

Serge Pajak (RITM, Université Paris-Saclay)

 

StreamOps

This research project aims at creating a novel approach, and tool to develop powerful algorithms capable of managing and mining data flows. 

It aspires to be at the interface of algorithmic, business and software aspects to offer both researchers and engineers a generic data stream processing and machine learning platform, at the forefront of algorithms through: easy integration of new algorithms, operational robustness, detection and information compression performance, and consideration of data confidentiality and problems related to real data.

Project directors: Marc Fischler (Hôpital Foch, UVSQ), Cédric Gouy-Pailler (CEA), Karine Zeitouni (UVSQ), Yehia Taher (UVSQ)

 

Michèle Sebag

Michèle Sebag has explored multiple approaches to AI. Graduated from the ENS, agrégé of mathematics, she began her professional career in the industry at Thomson before creating an AI consulting start-up. She then joined the CNRS, first at the Solid State Mechanics Laboratory at École polytechnique, then at Paris-Sud University at the Computer Research Laboratory. Among the feats of his teams: the development of an algorithm capable of beating human players in the game of go in 2009, nearly seven years before the AlphaGo DeepMind. Today, she works in particular on causal modeling, which she wants to use to reduce the biases of machine learning. Introducing causality into algorithms would, in particular, make the results of machine learning more explicable.

 

Laurence Devillers

Laurence Devillers is a full Professor of Computer Sciences and Artificial Intelligence at Sorbonne University/CNRS (LIMSI lab., Orsay) on Affective Robotics, Spoken dialog, Machine learning, and Ethics. She is the author of more than 150 scientific publications (h-index: 35). In 2017, she wrote the book «Des Robots et des Hommes : mythes, fantasmes et réalité» (Plon, 2017). She is involved in the French Commission on the Ethics of Research (CERNA), The IEEE Global Initiative for Ethical Considerations in the Design of Autonomous Systems (P7008 on nudging) and the HUB France AI.

 

Nicolas Anciaux

Nicolas Anciaux is research director at Inria, where he leads the PETRUS team. His areas of expertise are data management, trustworthy computing, privacy and data security. He is currently studying the design, implementation and evaluation of Personal Data Management Systems (PDMS), to help citizens manage their personal data under control. In particular, Nicolas focuses on secure processing of distributed queries using trusted hardware components nowadays integrated in user devices (PC, smarphone, IoT device). Nicolas is architect for the PlugDB platform, an embedded data management engine for IoT objects with large storage capacities (GBs), used today as a PDMS dedicated to home care. With legal researchers, he also leads the GDP-ERE initiative to co-construct technical-legal frameworks for personal data management for citizens. Nicolas is associate editor at the VLDB Journal. He is co-author of more than 50 conference and journal articles.

 

HistorIA

This project brings together researchers in history, computational social sciences and information visualization.

The goal is to develop and explore large historical databases by applying data mining methods supported by visualization, while implementing an iterative approach to the exploration process, based on users’ appropriation of the procedures, tools used, and the results of the analyses.

To this end, emphasis will be placed on the explainability  of the algorithms and on the progressive analysis of the data and the human-machine interaction.

 

Project directors: Jean-Daniel Fekete (Inria), Christophe Prieur (Télécom Paris)

 

GDP-ERE

Within the framework of Personal Cloud tools, depending on the architectures, it is the user who can be qualified as the controller of his data.

This research project focuses on the distribution of responsibility between the user and the supplier in Personal Cloud architectures.

Its objectives are the following: to analyze the impact of current Personal Cloud architectures on user responsibility and compare this analysis with the legislation and rules laid down by the RGPD; and to formulate legislative and technological recommendations in order to preserve the autonomy of the user.

 

Project directors : Nicolas Anciaux (Inria), Mélanie Clément-Fontaine (UVSQ), Philippe PUCHERAL, Guillaume Scerri (UVSQ, Inria), Célia Zolynski (University Paris 1 Panthéon-Sorbonne)

PEPER

How to make production, consumption and storage cooperate for a better use of renewable energies?

The PEPER project is studying and developing three concepts in order to create a balanced system for the efficient management of renewable energies: production, consumption, and storage. The objective is to collect data on the different actors of this network, and to use learning and deep reinforcement learning techniques to develop algorithms to predict the production and consumption of each actor, and then to allow cooperation between them.

 

Project directors: Hossam Afifi (Télécom Paris), Jordi Badosa (École polytechnique), Florence Ossart (CentraleSupélec)

 

Frédéric Pascal

Frédéric Pascal is a full Professor in the L2S laboratory at CentraleSupélec. From 2017, he is the head of the “Signals and Statistics” group of L2S. He is also the coordinator of the AI activities at CentraleSupélec and the chair holder of the Givaudan chair on data sciences. From 2017, he is in the Executive Board of the DATAIA Institute as the Program Coordinator.  Frédéric is also a member of the IEEE Signal Processing Society SAM technical committee, and he serves as Associate Editor for the IEEE Transactions on Signal Processing, for the EURASIP Journal on Advances in Signal Processing, and for Elsevier Signal Processing. His research interests are estimation, detection and classification for statistical signal processing and applications in radar and image processing. He is the author / coauthor of more than one hundred papers in the top journals and conferences in Signal Processing, Image Processing and Statistics.

 

Fabian Suchanek

Fabian M. Suchanek is a full professor at the Telecom Paris University in France. Fabian developed inter alia the YAGOknowledge base, one of the largest public general-purpose knowledge bases. This earned him a honorable mention of the SIGMOD dissertation award and the 10 year Test of Time Award of The Web Conference (WWW 2018). His interests include information extraction, automated reasoning, and knowledge bases. Fabian has published around 90 scientific articles, among others at ISWC, VLDB, SIGMOD, WWW, CIKM, ICDE, and SIGIR, and his work has been cited more than 10000 times.

 

Sarah Cohen-Boulakia

Sarah Cohen-Boulakia is a full Professor at the Laboratoire de Recherche en Informatique at Paris-Saclay University. She has been working for fifteen years with multi-disciplinary groups involving computer scientists and biologists of various domains. She spent two-years as a postdoctoral researcher at the University of Pennsylvania, USA and 18 months at the Institute of Computational Biology (IBC) of Montpellier in Inria groups. 
Locally she is member of the Center for Data Science steering committee.
S. Cohen-Boulakia’s research interests include provenance and design of scientific workflows, reproducibility of scientific experiments, integration, querying and ranking in the context of biological and biomedical databases. She is actively involved in the CNRS GDR MaDICS on data sciences and takes the lead of the GDR in January 2020.

 

Ioana Manolescu

Ioana Manolescu is a senior researcher at Inria and a part-time professor at École polytechnique. She is the lead of CEDAR, Inria project-team, focusing on rich data analytics at cloud scale. She is a member of the PVLDB Endowment Board of Trustees, and has been recently an Associate PVLDB editor, a president of the ACM SIGMOD PhD Award Committee, and a chair of the IEEE ICDE conference. Her main research interests are computational fact-checking and semi-structured data management. She has co-authored more than 150 articles in international journals and conferences, and contributed recently to a book on Web Data Management.

© DATAIA - www.dataia.eu

Création du site par RUBRIK C - www.rubrikc.com