EFLSemLex : une ressource lexicale graduée et désambiguïsée pour les apprenants de l’anglais en L2. Enrichir les ressources lexicales pour l’apprentissage des langues étrangères par la désambiguïsation lexicale
Files
Lannoo_28961600_2022.pdf
Open access - Adobe PDF
- 3.28 MB
Details
- Supervisors
- Faculty
- Degree label
- Abstract
- In this master’s thesis, we present a first version of EFLSemLex, a new resource for English as a foreign language. EFLSemLex includes frequency distributions for 14.662 words attested in expert-written textbook texts and online resources graded along the scale of the Common European Framework of Reference (CEFR). Fundamentally, the resource can help its users (EFL teachers, among others) to distinguish what kind of vocabulary should be comprehended by learners of English at a particular proficiency level. The particularity of EFLSemLex, as compared to the existing CEFR-graded resource for English (EFLLex – Dürlich & François, 2018), lies in the semantic aspect of the resource. Instead of identifying frequency distributions for word forms, EFLSemLex establishes frequency counts for word senses. The identification of word senses in the corpus is achieved by an algorithm of automatic word sense disambiguation. In parallel, this work offers a more general reflection on why resources aimed at learners of a foreign language should take senses into consideration, in addition to bare word forms.