||Managing construction knowledge in patterns:a neural network approach
||Hegazy T, Moselhi O, Fazio P
||Neural networks are AI-based computational tools with powerfulcapabilities of effective capturing and re-use of domain knowledge that areinherently implicit. This paper describes the modelling capabilities of neuralnetworks with respect to construction problems, emphasizing the advantagesassociated with their representation of knowledge in the form of patterns.Several aspects related to proper management of knowledge are addressed forthe purpose of developing practical and more reliable neural network modelsof complex construction problems. These aspects include: 1) problemstructuring and patterns formation; 2) knowledge acquisition and datavalidation; 3) preparation and transformation of acquired data; and 4) analysisand interpretation of network state of knowledge. Guidelines pertaining tothese aspects are provided along with considerations for modelling with noisydata and under high degree of uncertainty. The issues discussed are illustratedthrough a case study of a neural network for bidding decision support,developed based on knowledge acquired from contractors in Canada and theUS. The case study demonstrates neural network modelling and illustrates thebenefits gained through better management of acquired knowledge.
|Year of publication:
||neural networks; knowledge acquisition; construction; information technology;bidding strategy
Hegazy T, Moselhi O, Fazio P (1993).
Managing construction knowledge in patterns:a neural network approach. Mathur K S, Betts M P, Tham K W (editors); Management of information technology for construction; Singapore, August 1993,