User-driven Design of Ontology-based, Context-aware and Self-learning Continuous Care Applications Femke Ongenae and Filip De Turck Department of Information Technology (INTEC), Ghent University-IBBT Gaston Crommenlaan 8, bus 201, 9050 Ghent, Belgium Email: Femke.Ongenae@intec.ugent.be Abstract—A plethora of technologies is adopted in modern continuous care settings. Nowadays, the nurse is responsible for orchestrating all these devices to take advantage of the collected information. This causes caregivers to lose time, miss out on patient insights and lack a general overview of the situation. In this paper, an ontology-based and self-learning context-aware platform is presented. This platform combines pervasive and context-aware techniques with knowledge man- agement techniques to leverage the aforementioned problems and allows to easily build intelligent applications to support caregivers in their everyday activities. Moreover, a self-learning framework is proposed, which allows detecting trends in the usage of these applications and adapting them accordingly. However, the integration of context-aware technologies can prove to be difficult if domain experts are not involved in the development process. This increases the fear of the technology, makes the staff feel less in control and empowered and causes that the technologies are not tuned towards the work processes of the staff. Therefore, a participatory ontology engineering methodology is also presented which allows involving the stakeholders in each part of the development process of context-aware applications. I. I NTRODUCTION In recent years the complexity in nursing organizations has increased due to societal factors, e.g., the increase of the care unit size and specialized care and the lack of nurse staffing which requires a more efficient use of resources. A further increase of complexity is due to the high amount of new technologies that are being adopted, especially to support administrative tasks, data management and patient monitoring. The challenge today is that several devices have to be manually combined and consulted by the staff to take advantage of the information collected by all this technological equipment, even when carrying out one single task [1]. Due to the fact that the available data is not being integrated and aggregated, caregivers lose time, miss out on potential patient insights and lack a general overview of the situation. To cope with these problems, ambient-intelligent, perva- sive and context-aware techniques are often introduced. The common denominator of these techniques is that the technol- ogy will blend into the background of the environment and sensors and actuators will be able to sense and adapt our environment [2]. This implies an emerging demand for the integration and exploitation of the heterogeneous information available from all the devices such that the caregivers no longer play the role of the orchestrator between all these technologies. Moreover, this information integration allows building intelligent applications which exploit all this available data to support the caregivers in their everyday activities. To mediate the aforementioned problems, this paper presents an ontology-based context-aware platform, which allows easily developing pervasive applications. To easily build intelligent applications, the platform must be able to interpret the meaning and adequately filter the relevant information out of the huge amount of heterogeneous care data provided by the all the devices and sensors. Unorganized data is voluminous, but has no meaning on itself as it has no relationships or con- text. To transform the data to information, which is data that has been given meaning by defining relational connections, the platform employs ontologies [3]. An ontology is a semantic model that formally describes the concepts in a certain domain, their relationships and attributes. In this way, an ontology encourages re-use and integration. By managing the data about the current context in an ontology, intelligent algorithms can be defined that take advantage of this information to automate, optimize and personalize the continuous care of patients. Ontologies thus effectively separate the domain knowledge, which can be re-used across different applications, from the application logic, which can be written as rules on top of the ontology. However, the introduction of context awareness into com- mercial products is lagging behind what could be expected. This is due to inadequate techniques for personalization of the services, a lack of focus on the soft aspects of interaction, e.g., automated and personalized alerts, and the lack of tackling problems such as the fear for technology and the need of the users for control [4]. To ensure that technology and environ- ment melt into each other, the users should be involved in each step of the development cycle of context-aware applications. To achieve such a user-centered development process, this paper presents a participatory ontology engineering method- ology, which involves the users in each step of the ontology life-cycle. Finally, to make the developed context-aware applications more adaptable to future needs, this paper proposes a self- learning framework which allows detecting trends in the usage of these applications and adapting them accordingly. 978-3-901882-48-7 c  2012 IFIP