Augmentative and alternative communication and natural language processing: current research activities and prospects

Abstract

Historically, there has been little research into the use of natural language processing (NLP) within the context of electronic augmentative and alternative communication (AAC) systems. This is despite the fact that key aspects of AAC research are concerned with the treatment of natural language, and that communication aids appear to represent an ideal means of applying advanced NLP techniques. The lack of NLP research in relation to AAC is partially due to the tendency to focus NLP activities on solving particular problems from constructed examples, rather than the treatment of unrestricted language. Today, however, the face of NLP research has changed significantly, thanks to the increasing availability of and need to process larger corpora. This has prompted a quest for robust solutions to treat unrestricted text, which, in turn, has had two key results: (a) an influx of statistical techniques and (b) the emergence of comprehensive, language-related resources such as broad coverage electronic dictionaries. This paper describes current AAC research that uses NLP and comments on future research directions. Included is a brief survey of AAC systems and research prototypes involving NLP techniques, which is followed by an overview of resources emerging from NLP research that may be applicable to AAC.