On designing interactive systems that support creative problem solving in expert domains

Finding creative solutions to ill-structured problems is integral to the work in many expert domains. A common flaw of software tools that support this kind of work is to support mainly the detailed specification of a selected solution. To extend this support to the other processes of ill-structured problem-solving, I propose ten design principles, synthesized from results in diverse fields of research. These processes emphasize generating and comparing many potential solutions. To evaluate the principles' effectiveness, I built two prototypes; quantitative and qualitative results from evaluations demonstrate benefits, including faster task completion and the consideration of a wider variety of solutions. As there is disagreement within human-computer interaction on how to conduct such broad-scoped research, I introduce a generic framework modelled on the legal system and Thagard's explanatory coherence theory to structure this evidence into a compelling argument for the principles' wider adoption.