1. Skip to Menu
  2. Skip to Content
  3. Skip to Footer

Séminaires

 

Nos séminaires peuvent avoir lieu à l'EPHE (Bâtiment Le France : 190, avenue de France - 75013 Paris) ou à la Cité des Sciences et de l'Industrie (75019 Paris).


 

Parution de l’article Interpretation of approximate numerical expressions: Computational model and empirical study

 

L’article de Sébastien Lefort, Marie-Jeanne Lesot, Elisabetta Zibetti, Charles Tijus, Marcin Detyniecki, intitulé Interpretation of approximate numerical expressions: Computational model and empirical study, a été soumis à Elsevier en décembre 2016.

Ci-dessous, le résumé de l’article :

Approximate Numerical Expressions (ANEs) are linguistic expressions in- volving numerical values and referring to imprecise ranges of values, illustrated by examples such as “about 100”. In this paper, a general principle is proposed to interpret uncontextualised ANEs as intervals of denoted values. It is based on an empirically justified combination of characteristics of numerical values, both arithmetical and cognitive, and in particular, taking into account the cognitive salience of numbers. This general principle is instantiated in two computational models that can be extended so as to take into account the applicative context. An empirical study is conducted to assess the performances of the two models, comparing them to state-of-the-art methods, on real interpretations collected through an on-line questionnaire. Results validate the proposed characteristics used to build the models and show that they offer the best performances in es- timating the median interval chosen as representative of the collected intervals.

Parution de The Making of Smart Things dans le Journal of Science and Innovation

 

L’article de Charles Tijus, Marie Rougeauxet Javier Barcenilla, intitulé The Making of Smart Things, a paru dans le Journal of Science and Innovation (Vol. 6, No. 1, June 2016, 41-45).

Résumé de l’article :

Smart Thingsare objects that are connected (IOT) and dedicated for making people daily life simpler. Because Smart Things are taking decision for people and, for doing so, have to be adapted to their users, they are made of cognitive technologies that are technologies that include knowledge about human and about human cognition in order to process the data users are providing when interacting with Smart Things. The problem under investigation is how to design future Smart Things through prospective ergonomics. Derived from the action-based theory of categorization applied to contextual decision-making, the theory we present is based on deriving inferences from the multi-layers of linked properties (surface, structural, functional, procedural and behavioral properties) for the on-line cognitive computing of usages of Smart Things with sensors. Some examples are provided, emphasizing the role of the part-whole relational structure in designing future Smart Things.

Parution de l’article Automatic Extraction of Topics from Documents: Five Probabilistic Topic Model Tests dans International Journal of New Technology and Research (IJNTR)

 

L’article de Sandra Jhean-Larose, Nicolas Leveau, Guy Denhière et Ba-Linh Nguyen a paru en novembre dernier dans le International Journal of New Technology and Research (IJNTR).

Ci-joint le résumé de l’article :

In this paper, we test the capability of the Topicmodel to extract topics from documents (Griffiths & Steyvers, 2003, 2004 ; Griffiths, Steyvers & Tenenbaum, 2007). After presenting the mathematical aspects of the model and demonstrating its behavior on a small corpus, we attempt to falsify the model by manipulating (i) the size and similarities between the sub-corpora, (ii) the relative weight of sub-corpora, and (iii) the permeability to the scope and nature of contexts added to a fixed corpus. The model successfully passed our five tests, demonstrating that first, extracted topics were relevant and congruent to the content of the corpus, and second, that their probability appropriately reflected the relative weight of sub-corpora.

Projet ANR-FRQSC France-Québec - NUMEC

 

Le projet ANR-FRQSC France-Québec "NUMEC" a été sélectionné parmi 98 pré-projets soumis. Il s'intitule « L’apport  des outils numériques pour étudier l’écriture et améliorer son apprentissage en France et au Québec » et a pour objectif : (i) d’améliorer et d’articuler les outils numériques d’étude de l’écriture pour (ii) approfondir le développement de la graphomotricité chez les élèves français et québécois (dont les programmes scolaires diffèrent) et ses conséquences sur la réussite orthographique. Des retombées technologiques pour la recherche en Sciences humaines et sociales (numérisation de l’acte d’écriture et optimisation du traitement du signal associé) et sociétales (recommandations pour améliorer l’apprentissage de l’écriture en France et au Québec ; nouvelles modalités numériques d’évaluation des compétences en écriture) sont attendues. 

Hugues Delmas obtient un Award au Congrès EAPL

Hugues Delmas, doctorant au laboratoire CHArt et effectuant une thèse sous la direction d'Isabel Urdapilleta et de Charles Tijus intitulée "Détection du mensonge lors d'une surcharge cognitive par l'étude des cooccurrences d'indicateurs", a obtenu le premier prix de la meilleure conférence de PhD Student au Congrès EAPL (European Association of Psychology and Law).

Voir photo ci-contre (le deuxième en partant de la droite)