published in 2005 in UW Thesis archives
Architectural designers often use verbal-interaction metaphors when talking about the relationship between a designer and their drawings, as if they constituted partners in a conversation. Indeed, sketching or drawing often seems to serve a memory encoding function for the designer, helping to store rationale, purpose, and knowledge for later retrieval. Unfortunately, sketching mostly enables the designer to communicate to their drawing, but not vice-versa. Information retrieval from a drawing set tends to be either instantaneous (recovering a memory triggered by visual memory of the drawing configuration) or quite laborious (reconstructing knowledge amongst the human actors through file reference, further research, or secondary memory recall). This paper introduces a prototype drawing review tool that enhances the knowledge-recovery process required by the second case. Use of a customizable pen-based interface minimizes the cognitive load on the reviewer, while context-aware interpretation of sketch actions allows the reviewer to query the existing drawing (and the design knowledge base behind it) in a personal, natural, and fluid way. The reviewer's graphic queries are contextualized by the drawing under review, enriching the semantic meaning of the simple queries. Thus, sketching with intention on the existing drawing produces a query that varies the query according to drawing's properties. The system uncovers the possibilities of context-aware interface to query new meaning that drawing elements and reviewer's intention produce.
This thesis explores the relationship of construction documents to their related backing data and the process of document review. It is supposed that reviewers might benefit from simpler, smarter interfaces such as pen-and-gesture-based interaction. Further, the thesis explores the ways in which a graphical "selection" operation may be contextualized through consideration of drawing conventions: layer names, file names, etc.