Allergy Detector: um modelo ubíquo de detecção de riscos de alergia baseado na ciência de situação

Abstract

Advances in ubiquitous computing are enabling the emergence of opportunities in many areas, among them is the health area. In this area emerge many applications using ubiquitous computing for health care, called Ubiquitous Healthcare applications. According to survey conducted, have been found many models that enable ubiquitous healthcare to users, such as food planning, control intake of high-calorie foods, restaurant suggestions, daily monitoring of the diet and support in the selection menus as restrictions for safe diet. However, none of the models concerned provides support ubiquitous way for users who suffer from food allergies. Thus, this paper proposes to develop a ubiquitous model based on situation awareness, of risk detection intake of the eight major allergens (soy, egg, milk, wheat, fish, crustacean, trees nuts and peanuts) and their derivate, which causes about 90% of cases of all food allergies. The biggest model contribution to the scientific community consists of using the situation awareness for the specific purpose of supporting users in food allergy area. In addition, the model presents too an important contribution for society, supporting users who suffer from allergy to eight major allergens, presenting proteins contained in these foods or its derivatives, and that information obtained from the database hosted on the Union International Immunology Societies (WHO / IUIS) website. Was used the Endsley´s model as base to apply the situation awareness technique, which from the use of the profile and location contexts, added to the correlation of these two contexts, allows to perform the necessary inferences. And that from the correlation of these two contexts, the model is able to identify if there are dishes with allergens to the user's health. This correlation is only possible due to the ontology created, which stores all the information about the dishes and ingredients in these dishes the restaurant identified as well as the information of allergenic proteins contained in the eight major allergens. The proposed model had three evaluations, the first as an assessment by a case study, another on the application performance and a third evaluation on the usability of the model. For the case study were used the prototype created and confirmed the expectation that the situation awareness application, based on the Endsley´s model, would enable the model ubiquitously detect hazards to the user of the presence of allergens in food served in restaurants. While for performance evaluation, were collected the average response times for requests among the main services, and was measured CPU consumption during the sets applied requests, stating that the average response time increases linearly up a number of requests and thereafter presents an exponential behavior, and as the CPU consumption, it was found that the service used PaaSs platforms bit. As a final evaluation measured the usability of the model through field experiments with 10 volunteers, who testified that the model met all the measured constructs and from the results of statistical analysis, it can be confirmed that the proposed measurement model is consistent with the hypotheses identified on influence among constructs.