Automated and self-learning sentence generation methodology for Stammer patients by means of Natural Language Generation: Smart Speech Therapist for Stammer

摘要

The purpose of this research is to identify a suitable automated and self-learning treatment methodology for children suffering from stammer disorder. The first phase of the project involves interviews with Speech and Language therapists and doctors where I identify different types of treatment methodologies. I also used questionnaires to validate answers on the initial interviews. The final phase involves home visits where I use several assessments to confirm the child's diagnosis and examine the connection between manual and automated speech therapy process. By identifying the differences of results, I will validate the assumption that NLG (Natural Language Generation) with hidden markov model suits as a self-learning and automated sentence generation therapy methodology for stammer patients. This will allow for more individual consideration of children who suffer from stammer and may direct future research on the automated therapy methods for stammer.