Computação especulativa em inteligência ambiente

Abstract

A tomada de decisão pode ser vista um processo cognitivo pelo qual se escolhe um plano de ação entre vários outros, que pode ser baseado em vários cenários, ambientes, análises e fatores, para uma situação-problema. Todo este processo produz uma escolha/decisão final em que a “saída” pode ser uma opinião ou uma ação. Neste sentido a tomada de decisão é fundamental para a maioria das ações que realizamos diariamente de acordo com a situação e/ou problema com que nos deparamos. Para termos a percepção de todos os cenários possíveis numa tomada de decisão é necessário que tenhamos a possibilidade de especular sobre todos os casos, de forma a que possamos medir os prós e os contras consoante a ação que levarmos a cabo. Através deste trabalho pretende-se desenvolver uma plataforma de raciocínio especulativo, que seja capaz de acordo com uma determinada situação especular todos os cenários possíveis de modo a produzir decisões eficazes.

The decision making can be viewed as a cognitive process through which a plan of action is choose, among several others, for a problem-situation. This choice can be based on various scenarios, environments, analyzes and factors. This whole process produces a choice / final decision in which the "outcome" may be a belief or action. In this sense the decision making is critical for most actions we perform daily according to the situation and / or problem we face. In order to have the perception of all the possible scenarios in decision making is required the possibility to speculate on all cases, so that we can measure the pros and cons depending on the action that we carry out. Currently there are many systems that help us making decisions in order to increase our quality of life. The ambient intelligence (AmI) is designed to assist people in their needs and personal tasks. These environments require technology that captures information from the environment in order to be further processed to meet the needs of its users. The processing of the information captured by the technology in smart environments is an essential process. Often this information is not concrete and incomplete leading these systems to function incorrectly. In this sense it is necessary to use mechanisms that deal with uncertain or incomplete information in order to make effective decisions even when the information is unknown. In order to solve this problem it is introduced speculative execution in the scope of smart environments. In short, this work aims to develop a platform of speculative reasoning, which is able to meet the needs of users with visual disabilities, helping them on their dislocations. At the end of this work we obtained a system capable of guiding people with visual disabilities, assisting them and making their journeys safer and more autonomous.