Context-aware personalization environment for mobile computing

Dissertação para obtenção do Grau de Mestre em Engenharia Informática

Currently, we live in a world where the amount of on-line information vastly outstrips any individual’s capability to survey it. Filtering that information in order to obtain only useful and interesting information is a solution to this problem. The mobile computing area proposes to integrate computation in users’ daily activities in an unobtrusive way, in order to guarantee an improvement in their experience and quality of life. Furthermore, it is crucial to develop smaller and more intelligent devices to achieve this area’s goals, such as mobility and energy savings. This computing area reinforces the necessity to filter information towards personalization due to its humancentred paradigm. In order to attend to this personalization necessity, it is desired to have a solution that is able to learn the users preferences and needs, resulting in the generation of profiles that represent each style of interaction between a user and an application’s resources(e.g. buttons and menus). Those profiles can be obtained by using machine learning algorithms that use data derived from the user interaction with the application, combined with context data and explicit user preferences. This work proposes an environment with a generic context-aware personalization model and a machine learning module. It is provided the possibility to personalize an application, based on user profiles obtained from data, collected from implicit and explicit user interaction. Using a provided personalization API (Application Programming Interface) and other configuration modules, the environment was tested on LEY (Less energy Empowers You), a persuasive mobile-based serious game to help people understand domestic energy usage.