Text classification for children with dyslexia employing user modelling techniques

The problem of text-readability has received great attention in the literature. However, the classification of a text as readable is based solely in its linguistic complexity and does not take into account the skills of the intended reader. In this paper, we make a first attempt to study user-specific text readability. We focus on readers with dyslexia and documents written in English and Greek. Central to our approach is the notion of the user´s profile which carries information regarding the linguistic difficulties a user with dyslexia may experience. Based on the user´s profile, we develop heuristics for evaluating text´s readability for the specific user. The developed heuristics are incorporated in the text classification services of the iLearnRW1 project, aiming to facilitate the selection of appropriate/suitable reading resources for children with dyslexia.