GoCity: a context-aware adaptive Android application
Date
2012-12-14
Authors
Yang, Qian
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
GoCity is designed to provide city visitors with up-to-date and context-aware information while they are exploring a city using Android mobile phones. This thesis not only introduces the design and analysis of GoCity, but also discusses four problems in leveraging three concepts—context-awareness, self-adaptation, and usability—in current mobile application design. First, few contexts other than location and time have been used in actual mobile applications. Second, there is no clear classification of context information for mobile application design. Third, mobile application designers lack systematic mechanisms to address sensing and monitoring requirements under changing context situations. This is crucial for effective self-adaptation. Fourth, most mobile applications have low usability due to poor user interface (UI) design. The model proposed in this thesis addresses these issues by (i) supporting diverse context dimensions, (ii) monitoring context changes continuously and tailoring the application behavior according to these changes, and (iii) improving UI design using selected usability methods. In addition, this thesis proposes two classifications of context information for mobile applications: source-based classification—personal context, mobile device context, and environmental context; and property-based classification—static context and dynamic context. The combination of these two classifications helps determine the observed context and its polling rate—the rate at which the context is collected—effectively.
A distinctive feature of GoCity is that it supports two interaction modes—static mode and dynamic mode. In static mode, the application generates results only after the user sends the request to it. In other words, it does not actively generate results for users. In contrast, in the dynamic mode, the application continuously updates results even if the user does not send any request to it. The notion of an autonomic element (AE) is used for the dynamic mode to make GoCity self-adaptive. The polling rates on different contexts are also handled differently in the dynamic mode because of the differences among context properties. In addition, GoCity is composed of, but not limited to, four sub-applications. Each sub-application employs a variety of context information and can be implemented as an independent mobile application. Regarding usability, GoCity focuses on providing a simple and clear user interface as well as supporting user expectations for personalization.
An experiment which involves a person visiting the city of Victoria was conducted to evaluate GoCity. In this evaluation, three determining factors of usability were employed to qualitatively and quantitatively assess GoCity. In addition, the static mode and dynamic mode were evaluated separately.
Description
Keywords
context-aware, mobile application, self-adaptation