Modelo de aquisição de contexto afetivo baseado em autoanálise, classificação de tarefas e aspectos do estresse para a arquitetura ClinicSpace

• Ano de defesa: 2012
• Autor(a) principal: Fiorin, André
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• dARK ID: ark:/26339/00130000136d7
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria; BR; Ciência da Computação; UFSM; Programa de Pós-Graduação em Informática
• Palavras-chave em Português: Computação afetiva; Computação pervasiva; Contexto; Estresse; Atividades clínicas; Affective computing; Pervasive computing; Context; Stress; Linical activities; ClinicSpace; CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: http://repositorio.ufsm.br/handle/1/5411

The research in Pervasive Computing have directed their efforts to the modeling and development of programmable and interactive environments, able to assist users in their daily activities. The Systems Group of Mobile Computing (gMob) of Federal University of Santa Maria, developed the project ClinicSpace, a system to aid clinical tasks in hospitals based on Pervasive Computing technologies. On the other hand, the Affective Computing, which is a relatively new field of research in computer science, came up with the proposal to identify and synthesize human feelings in machinery, in order to make the interaction with computing devices more enjoyable and less frustrating. Given that pervasive systems work with context information to adapt their applications according to user needs, it becomes feasible through the Affective Computing, identify emotional characteristics of a person for this type of information can be used as context element, making the adjustment more precise applications pervasive. In this context, this research proposes a model of acquisition of affective context using techniques of inference of stress from psychological tools, emotional self-analysis and stressful classification of clinical tasks. Integrating the concepts of Affective Computing and Pervasive Computing, this model aims to classify the state of stress of the users of ClinicSpace and use it as an affective context, increasing the wealth of contextual information used in this system. To develop the proposed model, studies were performed on the Perceived Stress Scale (PSS), responsible for identifying stress in an individual, the use of AffectButtons to identify the user s emotional state, and the development of a questionnaire, applied to health care professionals to identify and classify the degree of stress of clinical activities. Based on these three approaches was possible to develop a classification model of stress for ClinicSpace users (doctors). To validate the proposal, was implemented a prototype of Affective Context Service for architecture ClinicSpace, which was tested in the Annas Dias Hospital (Ibirubá - RS). The results showed that the model proposed in this work is able to classify the state of stress for clinicians that this type of information can be used as an element of context in pervasive systems.