Simplify or help?

Abstract

We present a user study for two different automatic strategies that simplify text content for people with dyslexia. The strategies considered are the standard one (replacing a complex word with the most simpler synonym) and a new one that presents several synonyms for a complex word if the user requests them. We compare texts transformed by both strategies with the original text and to a gold standard manually built. The study was undertook by 96 participants, 47 with dyslexia plus a control group of 49 people without dyslexia. To show device independence, for the new strategy we used three different reading devices. Overall, participants with dyslexia found texts presented with the new strategy significantly more readable and comprehensible. To the best of our knowledge, this is the largest user study of its kind.