Towards a Diagnosis of Textual Difficulties for Children with Dyslexia

Towards a Diagnosis of Textual Difficulties for Children with Dyslexia Solen Quiniou, B´eatrice Daille Universit´e de Nantes, LS2N 2 rue de la Houssini`ere, 44322 Nantes Cedex 3, France {Solen.Quiniou, Beatrice.Daille}@univ-nantes.fr Abstract Children’s books are generally designed for children of a certain age group. For underage children or children with reading disorders, like dyslexia, there may be passages of the books that are difficult to understand. This can be due to words not known in the vocabulary of underage children, to words made of complex subparts (to pronounce, for example), or to the presence of anaphoras that have to be resolved by the children during the reading. In this paper, we present a study on diagnosing the difficulties appearing in French children’s books. We are more particularly interested on the difficulties coming from pronouns that can disrupt the story comprehension for children with dyslexia and we focus on the subject pronouns “il” and “elle” (corresponding to the pronoun “it”). We automatically identify the pleonastic pronouns (e.g., in “it’s raining”) and the pronominal anaphoras, as well as the referents of the pronominal anaphoras. We also detect difficult anaphoras that are more likely to lead to miscomprehension from the children: this is the first step to diagnose the textual difficulties of children’s books. We evaluate our approach on several French children’s books that were manually annotated by a speech therapist. Our first results show that we are able to detect half of the difficult anaphorical pronouns. Keywords: anaphora, dyslexia, children’s book, French 1. Introduction The democratization of books on digital tablets has allowed the design of new methods to support peo- ple with reading troubles. Children’s book publish- ers have thus proposed adaptations for young readers, with the use of specific typefaces, larger margins and text spaces, or refined illustrations to help them to bet- ter understand the content of the text. These adapta- tions have also proven their efficiency on Web browsers as they have allowed readers with dyslexia to better use their short-term memory on the text and on its mean- ing (Parilova et al., 2016). Research works have been mainly focused on using audio interfaces, as speech dic- tation, and screen readers to offer new features in dig- ital books addressed to readers with dyslexia (Sitbon et al., 2007) rather than on supporting the difficulties coming from the content of the text itself. Our work is complementary of the previous approaches as it only considers the content of the text. Indeed, we are interested in evaluating the difficulties that oc- cur in a text, in the form of textual ambiguities, and that can lead to difficulties in the comprehension of the story, especially for children with reading troubles like dyslexia. There are ambiguities at various levels: at the phonetic level such as the French word “fils” that can be translated as “son” or “threads” depending on the context and which is either a singular noun, or a plural noun (furthermore, its pronunciation is different in both cases); at the lexical …

Content truncated to 500 words