The paper by Antoine Amarilli and Louis Jachiet (from DIG) and Charles Paperman (U. Lille),
Dynamic Membership for Regular Languages, has won the best paper award at track B of the ICALP 2021 conference.

The paper by Antoine Amarilli (from DIG) and İsmail İlkan Ceylan (U. Oxford),
A Dichotomy for Homomorphism-Closed Queries on Probabilistic Graphs, has won the best paper award at the ICDT 2020 conference.

A new version of scikit-network is available!

This includes:

• accelerated code for massive graphs
• visualization in SVG format
• soft clustering
• soft classification
• fast embedding

Télécom Paris offers a full-time academic position as Maître de Conférences in the area of Artificial Intelligence, and in particular on techniques making results or decisions of AI explainable, starting September 2020.

More details here.

Pierre-Alexandre Murena‘s thesis: Minimum Complexity Principle for Knowledge Transfer in Artificial Learning (under the supervision of Antoine Cornuéjols and Jean-Louis Dessalles) got the 2nd prize of the best IMT thesis.

• Read on the IMT page (in French)
• See Pierre-Alexandre’s video presentation (in English)
• Read the thesis

Congratulations to Pierre-Alexandre!

The DIG team is opening an Associate Professor position in Scalable Artificial Intelligence at LTCI, Télécom ParisTech in Paris.

More information: here

University: Télécom ParisTech, https://telecom-paristech.fr/
Location: Palaiseau, near Paris, France
Position: Associate Professor (“Maître de conférences”), tenured permanent position
Application deadline: Friday, March 15, 2019
Starting date: September 2019
Team: Data Intelligence and Graphs (DIG, https://dig.telecom-paristech.fr/)

 

L’« IA » fait de plus en plus souvent la une des médias. Les mystérieux algorithmes de nos ordinateurs sont champions du monde d’échecs et de go, ils vont conduire nos voitures, traduire automatiquement en n’importe quelle langue, voire imiter nos modes de raisonnement. Hélas, ils ne savent même pas qu’ils sont intelligents.

Pour le dire plus clairement, ils ne savent rien. Tout ce que peuvent manifester les ordinateurs dotés des techniques les plus récentes d’IA est une intelligence qui ne comprend rien – du réflexe sans réflexion. Certains de nos mécanismes cognitifs, patiemment mis au point par l’évolution biologique, comme la recherche de la simplification et de la structure des phénomènes, sont encore hors de portée des machines, contraintes d’approcher au plus près de nos modes de raisonnement sans jamais les reproduire vraiment.

Le fantasme de la machine qui sait tout a donc de beaux jours devant lui, même si les progrès de l’IA posent avec toujours plus d’acuité la lancinante question de savoir si une véritable intelligence peut être produite par des circuits de silicium.

Jean-Louis Dessalles est enseignant-chercheur à Télécom ParisTech. Il utilise l’intelligence artificielle pour démonter les mécanismes de l’intelligence humaine, notamment en ce qui concerne le langage et le raisonnement.

Book Website

A workshop on Graph Learning will be held at LINCS on May 14h, 2018:

https://www.lincs.fr/workshop-on-graph-learning/

The objective of this workshop is to bring together people from industry and academia for presenting and discussing the most recent learning techniques based on graphs, from both theoretical and practical perspectives.

The workshop will cover the following aspects:

• Graph clustering
• Topic detection
• Recommender systems
• Graph-based classification
• Link prediction
• Graph alignment
• Social networks
• Dynamic graphs
• Graph signal processing

Speakers
Oana Balalau (Max-Planck Institute)
Alexis Benichoux (Deezer)
Pierre Borgnat (ENS Lyon)
Stephan Clémençon (Telecom ParisTech)
Vincent Cohen-Addad (CNRS / UPMC)
Matthias Grossglauser (EPFL)
Alexandre Hollocou (Inria)
Hervé Jegou (Facebook)
Renaud Lambiotte (University of Oxford)
Matthieu Latapy (CNRS / UPMC)
Dimitrios Milioris (Nokia Bell Labs)
Eric Siboni (Shift Technologies)
Michal Valko (Inria)

Organizers

Thomas Bonald (Telecom ParisTech)
Marc Lelarge (Inria)
Laurent Massoulié (Inria / Microsoft)

Our 2007 paper about the YAGO knowledge base has won the Test of Time Award of the World Wide Web Conference (WWW) 2018.

LTCI News Post

Fabian Suchanek, Professor at Télécom ParisTech, Gjergji Kasneci (SCHUFA Holding AG), and Gerhard Weikum (Max Planck Institute for Informatics) have been granted the 2018 Seoul Test of Time Award by the 27^th International World Wide Web Conference (also known as The Web Conference 2018) for their article entitled “YAGO: A Core of Semantic Knowledge Unifying WordNet and Wikipedia”, that was first published and presented in May 2007 at the 16^th Web Conference . This award recognizes the authors for their seminal work in creating the knowledge base, YAGO (Yet Another Great Ontology). Their awarded paper accounts for more than 2,468 cites to date. The Seoul Test of Time Award has been annually granted since 2014 to the author/authors of a paper presented at a previous World Wide Web conference that has, as the name suggests, stood the test of time.

YAGO, which is the result of a joint work with the Max Planck Institute, is nowadays one of the largest public knowledge bases created from Wikipedia, with more than 17 million entities and 150 million facts about these entities. Its source code was unveiled in 2017 (see our interview – in French). “YAGO was among the first projects to extract semantic knowledge at large scale from Wikipedia and is one of the pioneering contributors of the web of data, together with DBpedia”, Dame Wendy Hall, Professor at the Web Science Institute at the University of Southampton and Chair of the International World Wide Web Conference Committee (IW3C2), which supervises the Web Conference, said. “YAGO made available a large body of knowledge, samples of which were manually evaluated to show their very high accuracy. In addition to its contribution to knowledge content on the web, YAGO also played a seminal role in weaving a web of linked data by linking to other vocabularies including the DBpedia ontology and WordNet.”

The Web Conference is a prestigious yearly international conference, that was first organized in 1994, on the topic of the future directions of the world wide web. Three awards, in addition to the Seoul Test of Time, are granted. In 2018, Sergey Brin and Larry Page, who both founded Google, won the First Award for their paper entitled “The Anatomy of a Large-Scale Hypertextual Web Search Engine”, that was first presented at the 2015 Web Conference.

Article appeared at LTCI Chercheur en vue on  11 January 2018: https://ltci.telecom-paristech.fr/chercheur-en-vue-cat/fabian-suchanek/

Fabian Suchanek