Logging and managing places with PlaceWalker

Abstract

Managing and visualizing location data is an important part of many location-based service applica-tions. The most common way to express location information is to use numerical coordinates, e.g., latitudes and longitudes (e.g., “30.345 North, -65.443 West”). The key problem of this approach is that numerical co-ordinates are not so meaningful to people, and instead people often use diverse terms to express to their loca-tions on the basis of their own sense of place , such as “Starbucks with friends,” or “clinic in Home Plus,”. Recently, a numbered of place logging systems were proposed to enable automatic detection and logging of places that a user has visited. However, I observe that there is a significant gap in the granularity and meaning of detected places due to the mismatch of sense of places between humans and machines. Yet, none of the existing work in the ubiquitous computing literature has thoroughly investigated this issue. In this thesis, I first provide a comprehensive overview on how people perceive places by thoroughly reviewing related work spanning several disciplines such as humanistic geography, environmental psychology, spatial cognitive science, and socio-linguistics. From this review, I build a simple place management model that conceptualizes how people perceive and manage places. Given this model, I systematically analyze people’s diverse views on place via a two-week diary study, which confirms that, people’s sense of place varies from person to person and can be also changed over time. Finally, I design and implement the PlaceWalker system that, allows users to manage places from their own perspectives.