Automatic exercise generation to support macro-adaptivity in intelligent language tutoring systems

Abstract

Foreign language teaching achieves best learning outcomes when individual differences of learners are taken into account. While it is difficult for teachers to support internal differentiation in the classroom, digital tools can adaptively propose individual learning paths through activities so that students can practice with appropriately challenging exercises. But how can sufficiently varied, systematically parametrized exercises be provided to enable a system to match them to individual learner needs? We present an approach for high-variability exercise generation that transforms a single specification into a multitude of exercises varying in complexity. The approach is currently being evaluated in a randomized controlled study in regular German seventh grade English classes, facilitating a systematic empirical exploration of adaptive exercise complexity in relation to learning outcomes.

Exercise parameters influencing exercise difficulty

“…Since both classification and regression highlight the relevance of the generally applicable parameters, including linguistic complexity features such as token and character counts, we took a closer look at the impact of a range of linguistic complexity features on exercise difficulty. Didi's exercises of the learning targets conditionals and relative clauses were generated with the approach to systematic variability presented by Heck et al (2022), so that they contain learner data for exercises with identical textual material and varying only in a selection of controlled, syntactic features. These variations target the clause order, targeted clause and negation of clauses for conditionals, and clause order for relative clauses.…”