Pervasive Personal Information Spaces

Abstract

Each user’s electronic information-interaction uniquely matches their information behaviour, activities and work context. In the ubiquitous computing environment, this information-interaction and the underlying personal information is distributed across multiple personal devices. This thesis investigates the idea of Pervasive Personal Information Spaces for improving ubiquitous personal information-interaction. Pervasive Personal Information Spaces integrate information distributed across multiple personal devices to support anytime-anywhere access to an individual’s information. This information is then visualised through context-based, flexible views that are personalised through user activities, diverse annotations and spontaneous information associations. The Spaces model embodies the characteristics of Pervasive Personal Information Spaces, which emphasise integration of the user’s information space, automation and communication, and flexible views. The model forms the basis for InfoMesh, an example implementation developed for desktops, laptops and PDAs. The design of the system was supported by a tool developed during the research called activity snaps that captures realistic user activity information for aiding the design and evaluation of interactive systems. User evaluation of InfoMesh elicited a positive response from participants for the ideas underlying Pervasive Personal Information Spaces, especially for carrying out work naturally and visualising, interpreting and retrieving information according to personalised contexts, associations and annotations. The user studies supported the research hypothesis, revealing that context-based flexible views may indeed provide better contextual, ubiquitous access and visualisation of information than current-day systems.