Tiles — A model for classifying and using contextual information for context-aware applications

Abstract

Mobile information retrieval is characterized by the need to provide the right amount of relevant information for the right context so as to maximize the effectiveness of the limited screen size and processing capability of hand held devices. Context would typically include user profiles, user preferences, location, time of day, and others. To design better mobile information retrieval systems that typically draw information from various sources there is a need for a better understanding of these contextual properties that have an impact of design and delivery of such information. In this paper, we derive a set of 42 context properties and classify them using a model known as TILES (Temporal, Identity, Location, Environmental and Social). These properties were derived from literature in context-related studies, context-aware applications from the mobile tourism industry and augmented with focus group study findings. The TILES model, while having the highest applicability in the mobile tourism industry, can be adapted to other domains and serve as a framework to better aid the design and development of context-aware information retrieval systems in these domains.