“Modeling language learners’ knowledge:
What information can be inferred from learners’ free written texts?”

Sylvie Thouësny and Françoise Blin

School of Applied Language and Intercultural Studies
Dublin City University, Dublin, Ireland
sylvie.thouesny2@mail.dcu.ie, francoise.blin@dcu.ie

  Learner models store information about learners that enable intelligent tutoring systems to select the most appropriate pedagogical intervention for a particular student. Inferring a learner’s knowledge of grammatical forms from free written texts is however a complex endeavour. In particular, it requires the use of instruments that can help distinguish between errors (competence-dependent) and mistakes (performance-related). In this chapter, we explicate how we constructed such an instrument. We describe how a computer-assisted error encoding program (Markin) can be used in conjunction with a probabilistic part-of-speech tagging tool (TreeTagger), to tag a corpus of free texts produced by intermediate learners of French. Having computed the ratio of incorrect to correct forms, which can help us represent language performance at a given time, we then investigate whether low and high scores with regard to particular morpho-syntactic forms provide enough information to model learner knowledge. The findings of our preliminary analysis show that, on its own, the incorrect to correct forms ratio is often inadequate to help distinguish between errors and mistakes. However, the amount and type of assistance required by learners to successfully correct their written production provides valuable information on their actual language knowledge.  
Export citation - BibTex format
@Incollection {thouesny-blin:10,
 author={Sylvie Thouësny and Françoise Blin},
 title={Modeling language learners’ knowledge: What information can be inferred from learners’ free written texts?},
 booktitle={WorldCALL: International Perspectives on Computer-Assisted Language Learning},
 editor={Mike Levy and Françoise Blin and Claire Bradin Siskin and Osamu Takeuchi},
 address={New York},

Creative Commons License

Last revised on March 2019 - Copyright © 2007 onwards, Sylvie THOUËSNY, all rights reserved
All the content of this website is licensed under a Creative Commons Attribution-Noncommercial-No Derivative License