Analysis, ranking and prediction in pervasive computing trails

Many pervasive computing applications involve the recording of user interaction with physical and dig ital resources in the environment. Such records can be u sed to establish context histories that can be subsequentl y used for user behaviour analysis, pattern recognition, prediction, and the provision of context aware serv ices. In this paper we use trails as the principal data proc essing primitive for analysis and prediction. We define a trail as the sequence of recorded interactions with the perv asive computing space. Trails contain patterns of space u sage and they can be used for the provision of different services, space usage analysis or sociological info rmation of people using the environment simultaneously. Tra il analysis requires considerable storage and computat ional resources to discover such patterns. Moreover no si ngle method exists that identifies significant trails ba sed on different metrics for a variety of different pervas ive computing application. In this paper, we introduce a trail based analysis approach, an associated model for th e representation of trails and trail aggregates, and suitable data structures for efficient storage, filtering an d retrieval. Also, we propose several related algorithms and associated metrics for ranking and identifying sign ificant trails. We use these techniques in 2 different case studies to extract valuable information about the pervasive system environment usage and evaluate the summarizability and the predictive power of our model.