Cumincad : CUMINCAD Papers : Search Results

CumInCAD is a Cumulative Index about publications in Computer Aided Architectural Design
supported by the sibling associations ACADIA, CAADRIA, eCAADe, SIGraDi, ASCAAD and CAAD futures

PDF papers
References

Hits 1 to 5 of 5

Reformat results as: short short into frame detailed detailed into frame

_id	ee16
authors	Coyne, R.D., Newton, S. and Sudweeks, F.
year	1989
title	Modeling the Emergence of Schemas in Design Reasoning
source	Design Computing Unit, Department of Architectural and Design Science, University of Sydney, 1989. pp. 173-205. CADLINE has abstract only
summary	The authors explore how neural networks can be used to model important aspects of design reasoning: the way design involves memory; and a 'holistic' kind of reasoning by which designs appear to emerge from that memory. A simple neural network is constructed to demonstrate how information about schemas (in this case, room types) is stored implicitly after exposure to a number of examples of specific rooms. We then demonstrate how new room types emerge from this information. The paper includes a discussion of design, a discussion of schemas from a psycholinguistic perspective, a technical explanation of neural networks and the demonstration of an implemented examples
keywords	neural networks, modeling, design, reasoning, learning, knowledge acquisition, experimentation
series	CADline
email
last changed	2003/05/17 10:13

_id	68c8
authors	Flemming, U., Coyne, R. and Fenves, S. (et al.)
year	1994
title	SEED: A Software Environment to Support the Early Phases in Building Design
source	Proceeding of IKM '94, Weimar, Germany, pp. 5-10
summary	The SEED project intends to develop a software environment that supports the early phases in building design (Flemming et al., 1993). The goal is to provide support, in principle, for the preliminary design of buildings in all aspects that can gain from computer support. This includes using the computer not only for analysis and evaluation, but also more actively for the generation of designs, or more accurately, for the rapid generation of design representations. A major motivation for the development of SEED is to bring the results of two multi-generational research efforts focusing on `generative' design systems closer to practice: 1. LOOS/ABLOOS, a generative system for the synthesis of layouts of rectangles (Flemming et al., 1988; Flemming, 1989; Coyne and Flemming, 1990; Coyne, 1991); 2. GENESIS, a rule-based system that supports the generation of assemblies of 3-dimensional solids (Heisserman, 1991; Heisserman and Woodbury, 1993). The rapid generation of design representations can take advantage of special opportunities when it deals with a recurring building type, that is, a building type dealt with frequently by the users of the system. Design firms - from housing manufacturers to government agencies - accumulate considerable experience with recurring building types. But current CAD systems capture this experience and support its reuse only marginally. SEED intends to provide systematic support for the storing and retrieval of past solutions and their adaptation to similar problem situations. This motivation aligns aspects of SEED closely with current work in Artificial Intelligence that focuses on case-based design (see, for example, Kolodner, 1991; Domeshek and Kolodner, 1992; Hua et al., 1992).
series	other
email
last changed	2003/04/23 15:14

_id	e303
authors	Coyne, Richard D. and Newton, S.
year	1989
title	A Tutorial on Neural Networks and Expert Systems for Design
source	University of Sydney, 1989. pp. 321-337. CADLINE has abstract only
summary	This paper presents an overview of neural network (connectionist) systems and their potential contribution to computer-aided design. The authors discuss the appeal of neural networks and some of the problems. The major contribution to design is in the representation and manipulation of schemas. A neural network system can be 'taught' various examples (such as room descriptions). The system then apparently recognizes schemas (room types) and can produce novel but sensible combinations of descriptions constituting new types. A simple handworked example is presented, and the learning and reasoning mechanism is explained
keywords	representation, CAD, expert systems, design, neural networks
series	CADline
email
last changed	2003/05/17 10:13

_id	c3ec
authors	Coyne, Richard D. and Radford, Antony D.
year	1989
title	Knowledge-based Design Systems in Architecture : a Linguistic Perspective
source	Knowledge Based Systems in Architecture. Helsinki: Acta Polytechnica Scandinavica, 1989. pp. 27-36
summary	The separation of syntax and semantics in systems for the production of architectural designs is discussed. In producing designs we are less interested in mapping between buildings and their meaning than in mapping between an intended meaning and a design description of a kind that can be constructed, a mapping between semantic and syntactic realms. The thesis of this paper is that it is both feasible and operationally useful to separate the consideration of interpretation and syntactic generating in design. The authors first examine the possibility of operating in the syntactic realm, then the possibility of operating in the interpretative realm. Finally they examine how a design system might combine interpretative and syntactic systems
keywords	reasoning, design, CAD, architecture, knowledge base, semantics
series	CADline
email
last changed	2003/05/17 10:13

_id	8435
authors	Coyne, Richard D.
year	1990
title	Tools for Exploring Associative Reasoning in Design
source	The Electronic Design Studio: Architectural Knowledge and Media in the Computer Era [CAAD Futures ‘89 Conference Proceedings / ISBN 0-262-13254-0] Cambridge (Massachusetts / USA), 1989, pp. 91-106
summary	Two tools for storing and recalling information in computer systems are discussed and demonstrated in relation to design. The tools are hypermedia and neural networks. Each provides a valuable model for reasoning by the association of ideas.
series	CAAD Futures
email
last changed	2003/05/16 20:58

No more hits.