An application framework for mobile, context-aware trails
Citation:
Cormac Driver, 'An application framework for mobile, context-aware trails', [thesis], Trinity College (Dublin, Ireland). School of Computer Science & Statistics, 2007, pp 217Download Item:
Abstract:
Time management strategies for planning and scheduling activities increase the effectiveness of either personal or corporate time use. Supporting techniques are commonly based around the use of prioritised to-do lists. While the use of to-do lists for time management is beneficial, their static nature reduces their effectiveness in dynamic environments where users are mobile and activity properties can change over time. The prioritised ordering of a carefully considered, predefined to-do list can quickly become obsolete as its owner
begins addressing activities and unforeseen events occur. Mobile, context-aware computing is a computing paradigm in which small, portable devices such as Personal Digital Assistants (PDAs) and smart phones have access to information, known as context, about the situation in which they are being used and dynamically adapt application behaviour as appropriate. This paradigm facilitates automatic adaptation of a mobile user's schedule so that it accurately reflects the reality in which the user exists and maintains utility despite the occurrence of unforeseen events. Automatic, context-based schedule adaptation is at the core of a wide range of applications
for the mobile user who has a set of activities that may or should be carried out throughout the day at different locations. Implementing a mobile, context-aware activity scheduling application requires addressing
two common challenges. Firstly, an application must be capable of automatically
ordering a list of activities in an effective manner with respect to relevant context. Existing
approaches to mobile, context-based activity ordering are constrained in the number
of activities they can cope with or are server-based and subject to wireless network disconnection.
Secondly, an application must be capable of identifying when it is necessary
to reorder a list of activities to ensure that the list order maintains utility in the face of
context change. Techniques for identifying when it is necessary to reorder a list of activities are generally based on periodically assessing the ordering, resulting in the possibility
of an activity ordering becoming temporarily out of sync with the user's reality. To date,
mobile, context-based activity scheduling applications have typically been designed and
implemented in an application-specific manner - mostly as research prototypes. Consequently,
developers have had to repeatedly tackle the challenges inherent to this class of
application.
This thesis describes an application framework for the development of mobile, context aware
trails-based applications. A trail is a contextually scheduled collection of activities
and represents a generic model that can be used to satisfy the activity management
requirements of a wide range of context-based activity scheduling applications. The
framework supports developers by providing a generic, extensible implementation of the
trails model. Structure and behaviour common to mobile, context-aware trails-based
applications is provided, supporting context-based activity schedule composition (trail
generation), identification of whether or not schedule reordering is required following
context change (reconfiguration point identification) and subsequent automatic schedule
reordering as appropriate (trail reconfiguration).
The framework is evaluated through the development of three case study applications.
The case studies illustrate how the framework can be reused and extended to support the
development of a range of mobile, context-aware trails-based applications with differing
requirements. In addition, results of empirical experiments conducted to assess the responsiveness
of the trail generation implementation, the accuracy of the reconfiguration
point identification mechanism and human satisfaction with computer-generated trails
are presented.
Author: Driver, Cormac
Advisor:
Clarke, SiobhánQualification name:
Doctor of Philosophy (Ph.D.)Publisher:
Trinity College (Dublin, Ireland). School of Computer Science & StatisticsNote:
TARA (Trinity's Access to Research Archive) has a robust takedown policy. Please contact us if you have any concerns: rssadmin@tcd.ieType of material:
thesisCollections
Availability:
Full text availableMetadata
Show full item recordLicences: