Development of a goal-driven analysis for requirements definition in hypertext information systems supporting complex-problem solving
Albers, Michael Joel
MetadataShow full item record
When engaged in open-ended problem solving, the user must evaluate information from multiple sources. Unfortunately, people find it difficult to effectively search for and integrate multiple sources of information, requiring the system to provide the information in a manner which relates to the context of the problem. Also, rather than needing information in pre-defined ways, the viewing order and specific information requires changes with each problem. As a result, the methods used in conventional task analysis, which focus on defining the individual steps of a well-defined sequence, fail to provide good requirements for systems intended for supporting open-ended problem solving. Rather than focusing on individual steps, this dissertation develops a goal-driven analysis methodology based on defining and relating user's goals and information needs. Unlike a task-based analysis, the goal-driven analysis methodology revolves around uncovering the user's goals, the information needed to achieve those goals, and the contextual relationships between information elements. The analysis strives to uncover the major potential problem-solving paths and the information required to support following those paths, to provide the problem solver with varied routes to solving a specific problem. The unique feature of goal-driven analysis is that, throughout the methodology, it focuses on maintaining a connection between the user's goals, information needs, and problem context. This dissertation integrates the technical communication, cognitive psychology, and situation awareness literature, and explores the socio/cognitive aspects of information design as they relate to complex problem solving. It begins by arguing that effective information presentation requires a match between the user's mental model, the real-world context, and the factors which contribute to situation awareness. The dissertation then derives a four-step methodology: ethnography, interviews, scenario development, and group discussion, to develop a goal/information diagram which captures a graphical representation of the user's goals and information needs. The goal/information diagram then becomes the foundation for the analyst to use when developing system requirements. The dissertation also provides an extended example of how to perform a goal-driven analysis.