From a Cognitive Model Towards an Assistive and Augmentative Written Language

This paper presents a discussion about assistive and augmentative natural language processing designed for certain disabled persons unable to communicate. Several approaches have been proposed, according to abilities of the writer. Here we distinguish two cases in the writer’s capacities: the writer knows alphabetic writing, or (s)he does not know it. In the first case, the idea is to assist the writer by completing the words or the group of words which are initially written. In the second case, pictograms are used instead of characters, but it must be decided if these pictograms represent concepts or words in a new writing system. If the pictograms represent concepts, the produced text may not correspond exactly to the wishes of the writer; whereas when the pictograms represent words, the writer has to change his (her) mental approach to write the words that (s)he has chosen in another way. A new application of natural language processing If the compilation of increasingly sophisticated computer languages ensures that the machine will perform the operations described by a high level language, it is not the case for natural language processing. Automatic processing of natural languages rises complex problems of adjustment between the user and the machine at different levels, on which the machine does not "understand" yet the natural language as humans do: for instance, human-machine dialogue, search by keywords, querying databases, text mining, and automatic understanding of texts.. However, it is doubtful whether the understanding by the machine is necessary for many applications. If the machine Copyright © 2011, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved. is used as an augmentative capacity to partially compensate a user's disability of language, the problem may be conceptualized otherwise, and the possibility of a cognitive model can be very useful. Texts are generally stored as character strings without any semantic representation built by the machine. In some applications, texts are organized according to domain ontologies, but this method of organization is useful only for certain types of restricted treatment. The different approaches proposed for dealing with natural language show that the cognitive and semantic problem remains. Therefore, the target of the user must be known in order to adapt a mode of language processing that can answer the given problem: n-grams (Boissière et al. 2006) Finite State Automates (FSA), or more cognitive methods overlaying our linguistic process, have been studied. Taking as starting point a problem of disability relieved by assisted writing, we discuss the use of different approaches such as n-grams, FSA, glossary of phrases or cognitive methods. Context: severe difficulties to write We have described problems of pictographic palliation of communication disability in several articles. Here, we show how an adapted syntactic typing and a good cognitive design of pictograms are necessary to offer disabled persons a method that helps them to write a text. First, we must understand and know what characterizes the disability, then find the "toolbox" of NLP (Natural Language Processing) to provide the most appropriate assistance. (Abraham 2000a), (Abraham 2000b), (Abraham. 2007), (Abraham 2008). 132 Proceedings of the Twenty-Fourth International Florida Artificial Intelligence Research Society Conference