Computer interaction system to identify learning patterns and improve performance in children with autism spectrum disorders

Autism Spectrum Disorders (ASD) affects one in every 110 children. Medical and educational research have demonstrated that ASD children's social skills and adaptation can be much improved, provided that interventions are early and intensive enough. The advancement of computer technologies and their ubiquitous penetration in people's life make them widely available to support intensive sociocognitive rehabilitation. Additionally, computer interactions are a natural choice for people with autism who value lawful and "systematizing" tools. A number of computer-aided approaches have been developed, showing effectiveness and generalization, but little quantitative research was conducted to identify the critical factors of engaging and improving the child's interest and performance. This thesis designs an adaptive computer interaction system, called Ying, which detects learning patterns in children with ASD and explores the computer interactive possibilities. The system tailors its content based on periodic performance assessments that offer a more effective learning path for children with ASD.