This research is part of a Labex CominLabs : it is financed by a LABoratory of EXcellence which supports overall collaborative research in all scientific domains that use Information and Communication Technology. Our project relies on a set of experiments to which people with autism participate and, from these experiments, we aim to report a new way of caring these people. More than that, we aim to define recommendation tailored to each unique person relying on his or her own affinity. Then, these recommendations will be the fruit of co-working involving two complementary analysis from two different approaches.
The first approach from this research consists in the contribution of mathematical statistics for computer vision and data mining. In our work, it evaluates quantitatively the influence of affinity on the behaviors of ASD people. More precisely, it tries to extract behavioral characteristics shared by many autistic people in presence of their individual affinity, through a global study of attention indicators. Then, we want to study these indicators and their correlation with the viewing of an affinity (or of a neutral object that arises no particular interest), while carrying out the same study with neurotypical people. We hope to observe statistically significant differences that demonstrate, either an increased attentional engagement of people with autism in presence of their affinity, or, on the contrary, a gaze withdrawing due to a too strong emotional charge.
The second approach takes root in psychoanalysis which allows to consider each person from what makes him or her unique. In our work, we observed that affinity fulfills that function. An affinity is always proper to an autistic mind in particular. Then, for one autistic person, it is much more than merely a passion because it is the point on which the whole pattern of thought hinges. Without this affinity, anxiety arises. In other words, if the first approach provides a global analysis of the relation that join people with autism and their affinity, psychoanalysis allows us to add the idea of uniqueness that serves a case-by-case treatment by considering each person as an exception (as unique) : it works with the elective affinity specific to this autistic mind.
We, computer scientists and psychologists, have based together on the observation that it is very difficult for people with autism to engage their gaze (just as we see it is also difficult for them to engage their voice). Then, we may learn about how much a specified interest is invested by each autistic person if we happen to observe some statistically significant change in the manner they look (or avoid !) their affinity. Here, affinity becomes the lever to care ASD people, as long as we consider the way of using it proper to each one.
Putting these two analyses together may help us to provide new fundamental breakthroughs that are necessary today to the therapy of autism.
