Presentation of raw text to language learners is not enough to ensure learning. Thus, we present the Smart and Immersive Language Learning Environment (SMILLE), a system that uses Natural Language Processing (NLP) for enhancing grammatical information in texts chosen by a given user. The enhancements, carried out by means of text highlighting, are designed to draw the users’ attention to specific grammatical structures and thus help them to notice their occurrence in authentic contexts. To assess the quality of the enhancements, we carried out an evaluation of 48 structures in terms of precision in different text genres. This diversity approximates the contexts in which a language learner should immerge.
Zilio, L., Souza Wilkens, R., & Fairon, C. (2018). Enhancing grammatical structures in web-based texts. In Kate Borthwick, Linda Bradley, Sylvie Thouesny (ed.), CALL in a climate of change: adapting to turbulent global conditions. https://hdl.handle.net/2078.5/234030