Génération de feuilles d'exercices de géométrie à l'aide d'énoncés indexés automatiquament

Abstract : In order to generate an exercise page on a particular knowledge, exercises have to be indexed according to the knowledge useful to solve them. Hence, the main idea of this project is to use an automatic problem solver, Argos, to help us deal with this indexation. Argos aims not only at solving geometry problems, but also writes their proofs just as a teacher or a pupil would. Consequently, the meaningful data have to be extracted from the proofs found by the solver. Argos automatically generates rules from the properties given to it in a declarative way. We have linked the rules with geometric properties in order to deal with them, because the granularity of Argos' rules is too fine. These properties have been categorised using an ontology of the geometry theorems (about 75 theorems to categorise 150 properties). The useful theorems are stored in a database, so it is easy to build an exercise page on a particular subject (e.g. Pythagoras' theorem) by interrogating it. Hence the system we propose provides an assistance to the teaching of the curriculum or to the planning of activities in an Intelligent Tutoring System.
Document type :
Conference papers
Complete list of metadatas

https://edutice.archives-ouvertes.fr/edutice-00000141
Contributor : Florence Galand <>
Submitted on : Tuesday, November 4, 2003 - 3:21:02 PM
Last modification on : Tuesday, September 17, 2019 - 10:37:41 AM
Long-term archiving on : Monday, March 29, 2010 - 5:48:00 PM

Identifiers

  • HAL Id : edutice-00000141, version 1

Collections

Citation

Mathieu Hibou, Jean-Marc Labat, Jean-Pierre Spagnol. Génération de feuilles d'exercices de géométrie à l'aide d'énoncés indexés automatiquament. Environnements Informatiques pour l'Apprentissage Humain 2003, Apr 2003, Strasbourg, France. pp.247-258. ⟨edutice-00000141⟩

Share

Metrics

Record views

323

Files downloads

319