In education, as in architectural history, theory, and design, complete and thorough analyses of architectural bodies and objects are indispensable. These analyses cover many different aspects of the subject, e.g., physical and contextual attributes as well as geometric, functional, typological and organizational relations. When the results are computationally integrated, new views and arguments can be deduced from these that transcend the individual abstractions. Achieving such an integration computationally is hampered by the fact that different informational aspects of an architectural system require different representational languages. We propose an alternative approach to the quest for a truly universal, computational language or model. This approach relies on a computational framework where each model or abstraction is specified in its own appropriate representation, though with a common syntax. Representations can be combined and integrated, and relationships between the abstractions An extensive analysis of an architectural object or body leads to a model composed of abstractions, each reflecting on a different aspect. Though separately described through drawings, diagrams, and texts, these abstractions relate in many ways, most commonly through shared components. An integrated model that recognizes these relationships yields more than only the original abstractions. We present a methodology for achieving such a rich representation and touch upon the tools and techniques necessary to implement this methodology. As an application of this methodology, we describe an interactive educational system for the presentation of architectural analyses.
[1]
Richard Coyne.
Tools for exploring associative reasoning in design
,
1990
.
[2]
Yu-Tung Liu,et al.
A connectionist approach to shape recognition and transformation
,
1993
.
[3]
Ulrich Flemming,et al.
Syntactic structures in architecture: teaching composition with computer assistance
,
1990
.
[4]
Lefteri H. Tsoukalas,et al.
Fuzzy and neural approaches in engineering
,
1997
.
[5]
Rivka Oxman,et al.
The computability of architectural knowledge
,
1990
.
[6]
Ömer Akin,et al.
An electronic design assistance tool for case-based representation of designs
,
1997
.
[7]
Rudi Stouffs,et al.
Sorts: A Concept for Representational Flexibility
,
1997
.
[8]
Bob Carpenter,et al.
The logic of typed feature structures
,
1992
.