Approaching a Model for Context-Aware Applications based on the Semantic Sensor Network
ISBN 978-85-88783-11-9
Authors
1Cangrejo Aljure, L.D.; 2Delgado Fernandez, T.
1UNIVERSIDAD NACIONAL DE COLOMBIA Email: ldcangrejoa@unal.edu.co
2HIGHER POLYTECHNIC INSTITUTE JOSE ANTONIO ECHEVERRÍA (ISPJA Email: tdelgado@ind.cujae.edu.cu
Abstract
Context awareness is essential for services and applications that should adapt, in a flexible way, to a changing and dynamic environment. The relevance and significance of the information provided to a user through their mobile device, is determined by the knowledge of the user's contextual information, which involves several variables in addition to the location and can be captured through sensor networks. Management of contextual information includes gathering, representation, interpretation, reasoning and dissemination of contextual data, and has been addressed from different modeling approaches. However, the most promising approach nowadays, is the ontology based context modeling. The use of ontologies to context modeling is a valuable instrument for the formal and explicit specification of the entities in the user environment. Furthermore, ontologies constitute reference elements for geospatial information integration, reusability and interoperability (semantics) processes. These are important challenges in context awareness and pervasive computing paradigms. So far, studies of existing ontologies have allowed to identify a reusable ontological network at different levels: An upper-level, to model general level entities, using ontological resources such as DOLCE + DnS UltraLite Ontology and COSMO (COmmon Semantic MOdel). A context level including ontologies such as SOUPA (Standard Ontology for Ubiquitous and Pervasive Applications), Cobra-Ont (Context Broker Ontology) and some resources from mIO! Ontology Network. The third level is the user-level, used to manage user data and user preferences, which includes resources such as Who Am I and STOUP (Spatio Temporal Ontology of User Preference). From a spatial point of view, the analysis of LinkedGeoData (an effort to add a spatial dimension to the Web of Data, by using information collected by the OpenStreetMap (OSM) project) is being undertaken. By adding semantic annotations in terms of time, location, and thematic data into sensor data it is possible to create context-aware applications and facilitating advanced query and reasoning. The sensor information is a key source of contextual information for context-aware applications. However, the use of sensor data has inherent issues such as the increasing volume of data, heterogeneity of the measured data, formats and sensors in addition to the requirements of integration and interoperability. These issues were considered in the Semantic Sensor Network Ontology, which is part of the proposed model. This work is aimed to approach a model for Context-Aware applications in a semantic perspective with semantic annotation of sensor data by reusing the ontology network described above and the Semantic Sensor Network Ontology posed by W3C –SSN-XG. Additionally, some implementation issues regarding the model proposed are also discussed.
Keywords
Context awareness ; Semantic Sensor Network; Ontology