Extracting High-Level Information from Location Data: the W4 Diary Example

Services for mobile and pervasive computingshould extensively exploit contextual information both toadapt to user needs and to enable autonomic behavior. Tofulfill this idea it is important to provide two key tools: amodel supporting context-data representation and manipulation,and a set of algorithms relying on the model toperform application tasks. Following these lines, we firstdescribe the W4 context model showing how it canrepresent a simple yet effective framework to enableflexible and general-purpose management of contextualinformation. In particular, we show the model suitability indescribing user-centric situations, e.g., describing situationsin terms of where a user is located and what he is doing.Then, we illustrate a set of algorithms to semanticallyenrich W4 represented data and to extract relevantinformation from it. In particular, starting from W4 data,such algorithms are able to identify the places that matter tothe user and to describe them semantically. Overall, weshow how the context-model and the algorithms allow tocreate an high-level, semantic and context-aware diarybasedservice. This service meaningfully collects andclassifies the user whereabouts and the places that the uservisited