Combating educational failure through innovative training applications is essential, especially for people with special needs. ALECTOR attempts to meet this challenge: to leverage document accessibility by developing assistive reading technologies for children with dyslexia and poor-readers. For these children, text simplification might be a powerful and possibly the only way to leverage document accessibility. The idea is not to impoverish written language, but to propose adapted versions of original documents that convey the same meaning, as a result of automatic linguistic transformations. The scientific issues addressed in this project include (a) readability assessment (identifying the complex elements of a text that make it hard to understand and read at an optimal speed), (b) lexical simplification (replacing the complex lexical units detected with simpler equivalents, while dealing with polysemy and multiwords expressions), (c) syntactic simplification (replacing complex sentences with shorter ones, avoiding subordinations, passive voices, parenthetical explanations, etc.), and (d) discourse transformations preserving text cohesion (replacing pronouns with their antecedents). The main innovative aspect is that the final tool will be tailored to poor and dyslexic readers: text transformations will be based on theoretical findings about the reading process and further refined by specific adaptations leveraging feedback from the targeted audience. As one of the key innovative deliverables, ALECTOR will propose a web-based application where simplified corpora will be available to teachers and speech therapists. The project brings together four complementary partners with substantial competence and expertise both in natural language processing and language impaired readers (dyslexic children): the Laboratoire d’Informatique pour la Mécanique et les Sciences de l’Ingénieur (LIMSI, Université Paris XI), the Linguistique, Langues et Parole group (LiLPa, Université de Strasbourg), the Laboratoire de Psychologie Cognitive (LPC, Aix Marseille Université) and the Laboratoire d’Informatique Fondamentale (LIF, Aix Marseille Université), which leads the consortium. An asset of the project will be the collaboration with the CENTAL (centre for NLP in Louvain-la-Neuve, Belgium) having a strong expertise in readability and text simplification.