IntuiSketch, a pen-based tutoring system for anatomy sketch learning - Inria - Institut national de recherche en sciences et technologies du numérique Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

IntuiSketch, a pen-based tutoring system for anatomy sketch learning

Résumé

This paper presents the first works on IntuiSketch, a penbased intelligent tutoring system for anatomy courses in higher education. Pen-based tablets offer the possibility to have pen and touch interaction, which mimics the traditional pen and paper setting. The objective here is to combine online recognition techniques, that enable to interpret the sketches drawn by the students, with tutoring techniques, that model the domain knowledge. IntuiSketch is able to analyze the student drawings relatively to a problem defined by the teacher, and generate corrective feedback. The online recognition is based on the bidimensional grammar CD-CMG (Context Driven Constraint Multiset Grammar) which models the document structure, coupled with a fuzzy incremental classifier, which is able to learn from few examples. The tutoring system is based on constraint modeling, which enables to define domain and problem knowledge, and to analyse the student production relatively to the constraints that have to be satisfied to solve the problem. In this paper, we present a new architecture for anatomy sketch targeted intelligent tutoring system that combines different techniques. We also present a qualitative study of the feedback that our first system version is able to generate on a case study.
Fichier principal
Vignette du fichier
IGS.pdf (2.97 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04195461 , version 1 (04-09-2023)

Licence

Copyright (Tous droits réservés)

Identifiants

  • HAL Id : hal-04195461 , version 1

Citer

Islam Barchouch, Omar Krichen, Eric Anquetil, Nathalie Girard. IntuiSketch, a pen-based tutoring system for anatomy sketch learning. International Graphonomics Society 2023, Oct 2023, Evora, Portugal. ⟨hal-04195461⟩
15 Consultations
23 Téléchargements

Partager

Gmail Facebook X LinkedIn More