Language Technology for Automatic Control in eLearning Tools

The paper introduces the results of several projects (ALLES (Advanced Long distance Education System), ILLU (Internet-Lehr-Lernmodule für die Übersetzerund Dolmetscherausbildung) AUTOLEARN (AUTOmatic tutor for lifelong language LEARNing), and the soon starting COMENIUS project ICE3 ‘Integrating CALL in Early Education Environments’) addressed in [2], [3], [5], [6] whose objective was and is to create e-Learning sites for students of translation and second language learners realised as autonomous learning systems. ‘Autonomous learning’ means that the student can work with the system on the computer without any support by a human tutor. The innovation in all these projects is that intelligent (!) automatic correction is provided for students’ input. ‘Automatic correction’ could be seen on a trivial level on the basis of a simple matching of characters, e.g. as a gap filling exercise: ‘Fill in the right preposition’. This is actually the kind of IT-based language learning at the moment. Intelligent correction as provided in our systems goes far beyond that. It is meant to be the (automatic) detection of the kind of error that has been committed by the student and the issuing of a pedagogically useful diagnosis of the kind of error. This kind of correction is provided on different levels, orthographic, syntactic and even on a content level. Our automatic correction tools (AC tools) are based on sophisticated language technology.