Project description

A significant proportion of autistic preschoolers have a strong interest in tablets and smartphones, despite the fact that their verbal communication skills are not yet highly developed. With some of these children, it can be difficult to guide them in their choices. Researchers, practitioners and parents have frequently noted that these children spontaneously turn to applications or videos that include letters, numbers, shapes and colors.

Through this exposure, many children quickly attain a self-taught knowledge of written codes, as well as a vocabulary that pre-dates their development of speech several years later. However, two important factors can hinder the development of this ability. Firstly, the repetitive activities characteristic of autism, i.e., children prefer known and predictable activities that they explore exhaustively, may prevent children from discovering new material (e.g., watching the same video over and over again). Secondly, there may be limited availability of more varied and enriching environments.

The objective of this project is to develop a recommendation system based on machine learning, implemented as a mobile application. To do so, the intelligent system discreetly captures the interactions of an autistic child with a device when he/she is watching a video. Based on the information collected, the recommendation system will select and present children with new activities based on their interests and preferences in order to develop their ability to manipulate complex visual components, including language components.