Adaptive Ε-learning system for language learning: Architecture overview

Computer-based language learning involves employing computer technology for the purpose of learning natural languages. In such educational setups, instructional software becomes a tutor responsible for the construction of learners' linguistic knowledge and development of their language competences and skills. This is possible if the software has the ability to deliver appropriate pedagogical and expert competences a language teacher possesses, such as presenting learning materials at an appropriate level, setting up and adapting tasks and learning activities, identifying deficiencies in learners' knowledge or offering constructive feedback on their work and achievements. Introducing adaptive and intelligent features to a system is thus inevitable. The paper at hand describes the architecture of one such e-learning system currently being developed at our institution. The system follows the Common European Framework of Reference for Languages (CEFR) and has a twofold purpose: a) identifying the CEFR level of language competence by employing an adaptive knowledge validation procedure, and b) systematically supporting learning by guiding learners through the learning domain towards higher proficiency levels of the CEFR. In addition, its design is based on the socio-cognitive theory of second language acquisition and the notions of activity and communicativeness.