Skip to Main content Skip to Navigation
Conference papers

Amélioration de l'interprétabilité du diagnostic cognitif de l'apprenant par catégorisation des composantes de connaissance

Abstract : Cognitive diagnosis is used to model the learner knowledge from the learners’ traces on an ITS for adapted feedback or prediction of learner future state. By studying the performance of algorithms of knowl- edge tracing and factor analysis models on datasets restricted to either declarative or procedural knowledge components (KC), we show that factor analysis models outperform on declarative KCs while knowledge tracing ones outperform on procedural KCs, even if it is complicated to compare them. These results suggest the interest of categorizing the KCs by nature in cognitive diagnosis algorithms and that this categorization allows the enhancement of learner model interpretability.
Document type :
Conference papers
Complete list of metadata

https://hal.archives-ouvertes.fr/hal-03292665
Contributor : Marie Lefevre <>
Submitted on : Friday, July 23, 2021 - 9:43:05 AM
Last modification on : Sunday, August 22, 2021 - 2:46:06 PM

File

EIAH2021_Allegre_et_al.pdf
Publisher files allowed on an open archive

Identifiers

  • HAL Id : hal-03292665, version 1

Citation

Olivier Allègre, Amel Yessad, Vanda Luengo. Amélioration de l'interprétabilité du diagnostic cognitif de l'apprenant par catégorisation des composantes de connaissance. 10e Conférence sur les Environnements Informatiques pour l’Apprentissage Humain, Marie Lefevre, Christine Michel, Jun 2021, Fribourg / Virtual, Suisse. pp.34-45. ⟨hal-03292665⟩

Share

Metrics

Record views

61

Files downloads

33