Established in 2011, the INRIA Awards aim to promote the contributions and successes of those who advance the computer and mathematical sciences, thereby contributing to the development of the digital world. Organized by INRIA and its historical partners, the Académie des sciences and Dassault Systèmes, three prizes were awarded this year to Jean-Bernard Lasserre, theCONVECS team (Inria/LIG/CNRS/Université Grenoble Alpes/Grenoble INP) and Serena Villata.

The three categories of Inria scientific prizes

• The Inria – Academy of Sciences Grand Prize, worth €25,000, which rewards a scientist who has made an outstanding contribution to the field of computer and mathematical sciences. This person must be working in a French institution or be an emeritus of a French institution.
• The Inria – Académie des sciences – Dassault Systèmes Innovation Prize of €20,000, awarded jointly by the Académie des sciences, Dassault Systèmes and Inria, rewards a scientist or a team of scientists (consisting of a maximum of 6 people) who have been particularly active in the field of transfer and innovation in the computer and mathematical sciences. These persons, of any nationality and affiliation, must be working in a French institution or be emeritus of a French institution.
• The Inria – Académie des sciences young researcher prize, worth €20,000, awarded jointly by the Académie des sciences and Inria, rewards a scientist under forty years of age (year of birth 1981 or later, with a delay of one year for each child), of any nationality and affiliation, working in a French institution, who has made a major contribution to the field of computer and mathematical sciences through his or her research, transfer or innovation activities

The 2021 Winners

The Inria – Academy of Sciences Grand Prize was awarded to Jean-Bernard Lasserre for his research work, particularly in mathematical optimization. This is recognition of an exemplary scientific career, with major contributions and applications in many fields of science and engineering.

Jean-Bernard Lasserre is a mathematician, CNRS emeritus research director at LAAS-CNRS and at the Toulouse Mathematics Institute (IMT – CNRS/INSA Toulouse/UT1 Capitole/UT2J/UT3 Paul Sabatier), and holds the “Polynomial Optimization” chair at the ANITI Institute of the Toulouse Midi-Pyrénées Federal University

Founder of the “Moment-Sum of Squares” methodology (based on positivity certificates of real algebraic geometry) for optimization, he studies its many other applications in various fields of science. He is also interested in promoting the Christoffel function (well known in approximation) as a powerful and simple complementary tool for data analysis.

He has devoted much of his research life to mathematical optimization, and explains the principle behind it: How do you solve an optimization problem? It is about finding the minimum value of a mathematical function with a complex form. Imagine yourself as a hiker in a rugged mountain range: the summits are the maximum values of the function, the valleys the minimum values,” illustrates Jean-Bernard Lasserre. The hiker knows this from experience: he reaches a minimum value when, in all directions, the slope goes up… but is he sure that he has reached the deepest valley (and therefore the smallest minimum value)? In global optimization, mathematicians look for the “global” minimum value (the bottom of the deepest valley) and not a “local” minimum value (the bottom of another shallower valley).

He also acknowledges the importance of the collaborations he has been involved in:

“Throughout my career, I have been lucky enough to be in the right place at the right time and to work with very high quality researchers.”

The Inria-Dassaut Systèmes innovation prize was awarded to The CONVECS team (Inria/LIG/CNRS/Université Grenoble Alpes/Grenoble INP): Hubert Garavel, Frédéric Lang, Wendelin Serwe and Radu Mateescu.

These four researchers contribute to the development of the CADP toolbox, dedicated to the modeling and verification of parallel and distributed systems. The objective is to automatically detect design errors in these highly complex systems.

The CONVECS team (Inria/LIG/CNRS/Université Grenoble Alpes/Grenoble INP) has been working for 15 years on a project concerning the field of formal methods for modeling and verification of distributed systems with asynchronous parallelism. The team carries out work to model asynchronous systems and make formal methods more accessible, compilation techniques for these languages and verification algorithms for these systems. All these techniques give rise to tools that are coherently integrated in a toolbox called CADP. Hubert Garavel recalls:

“Early computers ran on a single processor, which executed line by line what was imposed on it by the program. But the increase in performance leads to having multiple processors and programs capable of running at the same time, which gives rise to thorny problems of synchronization and communication.”

The Inria-Academy of Science Young Researcher Prize is awarded to Serena Villata, CNRS research fellow in computer science at the I3S Laboratory (CNRS/Université Côte d’Azur) and member of the Wimmics team (Inria Sophia Antipolis-Méditerranée/I3S)

Serena Villata creates AI (artificial intelligence) tools capable of analyzing the logical structure of any text, from political speeches to Twitter messages. This is a promising avenue for combating online harassment and fake news. An AI specialist, her current work focuses on argumentation, with a focus on the automatic analysis of arguments in legal and medical texts, fallacious arguments in political debates and harmful content in social networks (hate speech, disinformation). She is interested in combining formal argumentation, based on critical reasoning, with human argumentation.

Her journey in three steps

• His early work on formal argumentation theory, in 2007 – “It was a topic with a fundamental character, but it could give rise to multiple real-life applications: most exchanges between individuals are based on arguments. But at the time, these applications seemed futuristic and very ambitious…”
• Her beginnings in natural language argumentation, in 2012 – “I went from computer science theory to concrete and everyday life: automatically extracting arguments from messages on a forum, from a political debate, from a law text… Everything was to be built and I was one of the pioneers; it was the right time to get started.”
• Her projects on societal topics, from 2018 – “My research started to move towards the prevention of online hate, propaganda, misinformation… It’s important for me to feel that my work has a real social utility.”

“The big challenge of the moment in my field is to design a chatbot capable of dialoguing in natural language with users, for example schoolchildren, to help them develop their critical thinking. It could explain a subject, illustrate it with examples, oppose counter-arguments to the fake news it generates, describe the devastating effects of the propagation of these fake news, etc.

This is a very ambitious goal: the chatbot sociologists, etc. “should be able to automatically deal with any topic, without being prepared, by adapting to the interlocutor’s knowledge.Moreover, such a tool should not be misused for disinformation campaigns. Its design must be multidisciplinary, and involve legal experts, sociologists, etc. ….”

Translated from Retour sur les trois lauréats des Prix Inria scientifiques 2021 : Jean-Bernard Lasserre, l’équipe CONVECS et Serena Villata