Middleware and Context Inference Techniques from DataStreams for the Development of ContextAware Services using Mobile Devices
The ubiquity of the mobile communication devices, such as smartphones, allowed the emergence of context aware applications and services, which proactively respond to activities or specific situations of each user. The context information, that is, the specific state in which each user finds himself, allows the communication providers the development, and consequently the offer, of new services of added value to a great range of applications, such as social networks, advertising, navigation or leisure. Of growing importance are services and applications, related to the health sector, that depend on the precise detection of each user’s physical activity, whether in specific cases or along several days or weeks. Using this information, it is possible to discover and analyze the physical activity patterns and, for example, help individuals to have healthier life styles. The proposed project approaches the challenges and the techniques to develop context aware applications in three components: - Firstly, it proposes to research and develop activity detection techniques focused on low energy and high precision. - Secondly, it will explore context aggregation approaches and algorithms based on statistic classification, in order to discover activity patterns. - Lastly, since the mobile systems interface will definitely evolve, the project will develop a middleware and a programming environment, specific to the fast prototyping of context sensitive applications. As part of the proposed objectives, the project will evaluate the techniques using a real prototype, which consists on smartphones and a set of sensors adapted to a personal vest, inspired on the prototype system previously developed. The middleware and the I&D techniques used during the project will be evaluated on an industrial context by Altice Labs, through a context sensitive and innovative application for social networks. This validation will allow the project team to estimate the viability of this approach on a real context and, in last resort, evaluate the possibility of a technology transfer.