Abstract for HONS 11/15
A Web-Based Framework for Personalised Language Rehabilitation
Roseanna Grundy
Department of Computer Science and Software Engineering
University of Canterbury
Abstract
This research project presents the research and development of web-based framework for simple
language rehabilitation software that personalises to patient needs. This prototype uses a simple
language task to demonstrate the viability of complete system that rehabilitates communication for
people with Aphasia (PWA). Current rehabilitation techniques are repetitive and drill-based with
little personalisation. Personalisation has been shown to enhance learning and maintain motivation
in educational systems. We implement these same techniques in the application framework with the
aim of enhancing quality of therapy and motivation in aphasic people. PWA can use the software
independently and still receive a degree of personalised therapy without a human speech and
language pathologist. Intelligent and adaptive software in this space does not currently exist. The
framework uses Intelligent Tutoring System (ITS) techniques to achieve this goal. The ASPIRE ITS
authoring tool was used to author a simple language tutor whose components would be integrated
into the prototype framework. The domain ontology was constructed, task sets created, constraints
generated, and feedback messages customised. This project found that the framework was viable,
but that some functionality provided by ASPIRE may be insufficient. The components of the
framework are designed to be modular, this way the constraints or feedback could be extended or
replaced to meet future requirements.