AN AFFECTIVE INTELLIGENT TUTORING SYSTEM IN THE SPECIAL EDUCATION OF INDIVIDUALS WITH AUTISM

Autism spectrum disorder (ASD) is a neurological disorder affecting the way in which the brain processes information. It can affect all aspects of a person’s development. Autism is characterized by impairments in learning and communication, in social interaction, in imaginative ability, as well as, in repetitive and restricted patterns of behavior (DSM-IV. American Psychiatric Association, 2000). This research contributes to the advancement of Intelligent Tutoring Systems by proposing an affective intelligent tutoring system to be used in the field of specialized education in order to overcome the lack of individualized intervention as pertains to special education for individuals with learning disabilities such as autism. The Integrated Specialized Learning Application (ISLA) is an Affective Intelligent Tutoring System (AITS) that evolves along with the learner’s needs. ISLA is unique and its contribution entails the computational model of accompaniment for helping autistic children manage their emotions by analyzing the learning trace during a mathematical activity such as addition. Its virtual pedagogical agent called Jessie calibrates the emotional state of the autistic learner by providing encouragement and support in real-time with feedback messages to help the ASD student cope with difficulties as he/she progresses during the activity at hand. We have conducted a study using a prototype of ISLA consisting of two main experiments that implemented Jessie as a virtual pedagogical agent based on our accompaniment model. The main experiments involved twelve participants with high functioning autism, among which one group of six students interacted with the pedagogical agent Jessie which takes into account their progression, while the control group of six participants interacted without the pedagogical agent. The findings of this research indicated that the majority of participants in the group without Jessie did not complete the quiz. The majority of participants did not succeed, either, on the quiz without the use of a pedagogical agent providing support and encouraging motivation during the mathematical activity. On the other hand, the results obtained with the test group revealed that the majority of participants benefited from the personalization and support provided by the pedagogical agent Jessie, which aimed at helping the autistic student to become self-regulated by calibrating his/her emotions and by encouraging motivation during mathematical learning. Emotional reactions using FaceReader technologies, observations as well as verbalizations were compared during the interaction in both groups and showed that participants using Jessie were less angry, less anxious, less frustrated, and less disengaged.