Eric A. Domeshek

A Case-Based Design Aid for Architecture

This paper summarizes the current status of a project to construct a design aiding system for architects. The Archie-II system is an application of case-based reasoning techniques to the task of assisting human designers. The focus on design aiding, the choice of case-based techniques, and the resulting specification of a case browsing system are reviewed and justified in the first section. The balance of the paper then focuses on the ways in which design cases can be carved up for presentation to designers and how the resulting pieces can be indexed and organized so as to make them available at appropriate times in the design process.

Using the points of large cases

We believe that Case-Based Reasoning (CBR) will prove applicable to design, at least in part because we have seen designers making extensive use of past cases. Construction of useful systems, however, requires resolution of many open issues. In this paper we consider three issues in particular: 1) What sort of content should be captured in a design case? 2) How should the content of a complex case be segmented into chunks for use? 3) How should the resulting chunks be indexed for retrieval? These are among the issues we are seeking to address through construction of Archie, a case-based aid for conceptual design in architecture. In addition to our approaches to these issues, this paper also discusses our strategic choice to build a design aiding system as opposed to a system that generates designs on its own.

The Design of a Tool Kit for Case-Based Design AIDS

This paper extends the discussion of Case-Based Design Aids (CBDAs) presented at the 1992 AI in Design Conference (Domeshek and Kolodner, 1992). After reviewing the CBDA concept and its motivation, we discuss progress in implementing such systems, and focus on the generalization of our original CBDA (first developed to support architects with the conceptual design of buildings) into a tool kit applicable to a wide range of design domains. Experience in gathering and organizing case materials, and consideration of more use-scenarios have led us to reorganize and refine many aspects of the original proposal. Here we report our experience building CBDAs for architecture and for jet aircraft subsystem design using the CBDA tool kit Design-MUSE.

A case-based design aid for conceptual design of aircraft subsystems

This paper describes MIDAS (a memory for initial design of aircraft subsystems), a system that applies insights and techniques from case-based reasoning (CBR) to aid engineers in the design of utility subsystems early in the development of a new aircraft concept. Our goal is to demonstrate the usefulness and practicality of a particular approach to building a corporate design memory. MIDAS is an instance of a general class of systems we call Case-Based Design Aids (CBDAs). A CBDA provides a designer with convenient access to multimedia presentations that highlight the outstanding good and bad points of previous designs. MIDAS was developed as a joint project of the Georgia Tech AI Lab and Lockheed Aeronautical Systems Companys Advanced Design Division. It is the first CBDA to be built largely by domain experts; the AI team primarily provided an (evolving) tool kit, and advice.<<ETX>>

Understanding the Creative Mind: a review of Margaret Boden's creative mind☆

Summary and conclusions Inference and the control of inference, knowledge representation and representational change: these are the main interrelated pieces of the creativity puzzle. Each relies heavily on episodic and semantic memory. Together, they fit into a model of reasoning that is recognizable as (but looser than) case-based reasoning. A creative individual is one in whom these factors combine to form a search space-a repertoire of thoughts-that is different from the usual and contains many creative ideas waiting to be constructed. Of course, the search space can only be explored in the context of a task or problem and a specific situation; thus, the repertoire is delined pragmatically, and serendipity (as Boden points out) plays an important role. In a specific individual, more creative thoughts will likely result when these pieces come together in a novel way to yield an unexplored and unexpected path through the search space. Creativity, as Boden points out, is not an all-or-none phenomenon. Every new thought is creative to some extent. Every new thought results from those same processes that, on occasion, produce results we value as creative. The more the search space is varied in a given context (through representational change, novel inferential methods, or strategic control heuristics), the more creative the resulting thoughts are likely to be. Over time, an individual may become more expert as he or she acquires (or reformulates) knowledge, reasoning strategies, and methodologies that change the search space or how it is explored. The framework we have sketched here is broadly compatible with Boden’s, but is more specific in its suggestions for integrating multiple types of interacting and interactive processes in a task context. In accounting for creativity, we emphasize issues of control and the role of experience (or cases). By focusing on how mental activity is directed towards a task in some situation, we ensure that the resulting theory addresses pragmatic issues in thinking and control of thinking. As Boden would require, our approach is computational. We believe, in fact, that the greatest contribution of The