PhD Dissertation Defense Presentation-School of Architecture - Carnegie Mellon University

Monday, September 13, 2010

PhD Dissertation Defense Presentation

Yeonjoo Oh, PhD Candidate, Computational Design

DATE:  Sept. 13, 2010
TIME:  2:00PM
LOC:   MMC 107 (Dean's Conference Room)

Abstract : Toward A Theory of Design Critiquing: The Furniture Design Critic Program

This thesis provides a computer-based critiquing system, the Furniture Design Critic. Although critiquing is pedagogically essential in design education, we lack a clear and systematic understanding of design
critiquing. For example, experienced studio teachers offer critiques depending on their intuitions and their own academic experiences. Also, in the field of design little is written about design critiquing, specifically, how to develop critiquing strategies.

In this PhD research, I first developed a theoretical framework of design critiquing practice through a literature survey. Based on this framework, I implemented the Furniture Design Critic program by adopting the typical system architecture of constraint-based tutors, because the theoretical framework cannot offer enough details about design critiquing, specifically, how to select critiquing methods in the consideration of critiquing conditions. This program offers tools in the forms of (a) an experiment for exploring and investigating alternative critiquing strategies; (b) a computational model to describe design critiquing and to demonstrate plausible ways to decide on critiquing methods; and (c) a computational environment where I can demonstrate how to make an inference about critiquing conditions such as a designers’ knowledge levels and select a certain set of critiquing methods.

My dissertation contributes to discussions in the fields of design, design education and computer-based systems regarding design critiquing. This Furniture Design Critic program developed in this work provides a foundation to describe and articulate critiquing strategies, specifically, how to select critiquing methods under which conditions. The program demonstrates a model of design critiquing; it makes an inference about critiquing conditions such as how much a designer knows, what weaknesses and strengths the designer has, what methods are effective to the designer, and an interaction history between the program and the designer; and it then selects a set of critiquing methods. This program also advances knowledge in the fields of computer-based systems including critiquing systems and intelligent tutoring systems by demonstrating how to make an inference about and consider critiquing conditions, and select a particular set of
critiquing methods based on these conditions.