Automated Capture of Design Knowledge from Design Process

Designers depend on computer as the main workbench to deal with their complex product design tasks today. So a mass of design process related digital data can be obtained. How to discovery design rationale knowledge from primary design process data is still an open problem. This paper describes an experimental system, the design history construction framework. In the framework, clustering techniques are used to aggregate design events into abstract summaries of tasks, and plan recognition algorithms are used to identify key episodes of activity.