Moving on from Location and Time to a Richer Understanding of Context in Ubiquitous Computing

Developing context-aware computing systems, such as emotion-adaptive services, require that the full range of relevant context variables is thoroughly identified. Various context meta-models systematically structuring the context world have been proposed recently for conceptualizing context in the requirements engineering phase (e.g. Bauer, 2011; Zainol & Nakata, 2010). Despite this guidance, only a few different and conveniently measurable context variables including time, location and device context are applied in real-world applications and research. In this paper we investigate the comprehensiveness of existing context meta-models, which is necessary for designing rich context-aware systems considering the full range of relevant context. Our analysis of context variables extracted from the "IEEE Pervasive Computing Magazine" (n = 297 articles) reveals that no context meta-model is comprehensive. Embedded in a modified process for conceptualizing context, we propose an enhanced context meta-model based on the model of Bauer (2011). Our model takes into account that only service domain-specific and generic context together can paint the complete context picture. 50 interchangeable domain-specific variable modules as well as generic context related to technology (IT, virtual world) and movement are additionally considered. The model has been empirically evaluated drawing on information systems design theory and qualitative content analysis.