Most previous research focused on designing Web applications on one or two aspects of Web applications development. We can classify existing studies into three different approaches based on their functionality: methodologies focusing on static Web sit ...
Most previous research focused on designing Web applications on one or two aspects of Web applications development. We can classify existing studies into three different approaches based on their functionality: methodologies focusing on static Web site design, approaches that have a focal point on the functional aspect, and studies that describe a higher-level integration of an e-commerce application with existing non-Web applications. However, little research has been done on a comprehensive design method. Thus, we propose a Web application design methodology based on information clustering, which is founded on several classifications.
We attempt to address three research issues in order to achieve our research objective.
Issue 1: Identifying important issues addressed by existing methodologies
The previous studies about Web application development indicate some useful design principles. First of all, a Web application should be designed in a way that reduces cognitive overhead, which is primarily related to user disorientation and user-interface adjustments. Second, to provide users with an effective navigation, a Web application should be characterized by higher local coherence, higher global coherence; and effective navigational facilities, that is, providing support for navigation with respect to direction (breadth) and distance (depth) within a cluster as well as across clusters.
Issue 2: Developing design constructs for the proposed methodology
This study proposes four critical factors that allow us to derive useful design constructs
(1) Web technology is rapidly advancing to meet the business computing need that demands more interactive and dynamic delivery of information. In addition, the delivery process involves one or more layers of a Web application, that is, client-side and server-side.
(2) Web applications are often integrated with existing legacy business applications. The objective is to leverage and extend existing critical business systems directly to customers, employees, suppliers, and distributors via the Web to improve time to market and reduce the cost of development and deployment. This suggests that we take into account the existing applications as reusable components during the course of Web application design.
(3) The dynamic nature of Web applications calls for a content management through supporting systems architecture. Content management dealing with changes seems to be equally important as developing an application. It is believed that a Web application should always be "under construction." In other words, the evolutionary aspect of Web applications is manifested by ever-changing structures, presentations, and contents themselves. Our proposed concept, compendia. should make it easier to update, add, or delete pages and links without serious "ripple effects."
(4) Webpages are connected through "links." The hyperlinks between information clusters based on natural flow of logic and rhythm of content are mostly associated based on their implicit semantic relationships. For example, user login Webpage contains a link that invokes a form page, and the form page might be connected to a database to store or retrieve the user input data. We, therefore, proposed nine different semantic link types.
Issue 3: Propose a framework of context-based Web application design methodology
We propose a context-based Web application design methodology that encompasses the design environment changes. The methodology is founded on several classification schemes described above (issue 2). Like conventional system development methodologies, it consists of three steps: requirements determination, analysis, and design. These steps are, then, further refined into 7 phases: organizing themes, determining access scope, shaping higher-level compendia, compendium refinement, elaborating compendium diagrams, developing link data dictionary.