Introduction

People with severe motor disabilities, such as muscular dystrophy, cerebral palsy and musculoskeletal disorders, often have difficulty communicating in written and/or oral form. In general, they are required to use an Augmentative and Alternative Communication (AAC). Examples of AAC include single-switch scanning using a virtual keyboard, soft keyboard used with adapted input device, etc.  In most cases, users will have training and need some practice to learn, use and benefit of an AAC. One of the main problems of AAC is that users interact with an adapted pointing/selection device. The use of a single pointer slows down text input. This requires repetitive movements that increases physical and cognitive load for the users, resulting in fatigue, errors, low acceptability.

To solve these problems, several solutions have been explored. One possibility takes the advantage of letter co-occurrence or word statistics in language in order to suggest a static or dynamic reorganization of the keyboard or other dynamic improvements. Another well-known technique is to provide a word completion list, which reduces the number of actions needed to enter a word or sentence. This requires advanced algorithms for data analysis and machine learning to predict text, and have great potential applications.

Challenges and call for contribution

This Special Thematic Session (STS) will discuss recent experience and contributions on work for enhancing AAC with a prediction system. This Special Thematic Session will consider all aspects of research on AAC and its interaction with prediction system, including (but not limited to):

• Measures and models
• HCI Design
• Prototype & tools
• Evaluation & User experience

Chairs

• 
  Mathieu Raynal, University of Toulouse
• Georges Badr, Antonine University