Data mining for intelligent structure form selection based on association rules from a high rise case base

Abstract

This paper presents a uniform model for high-rise structure design information and a case base containing 1008 high-rise buildings around the world. A case management system has been implemented with functions of data recording, modification, deletion, inquiry, statistical analysis and knowledge discovery. We propose a data-mining process of mining quantitative association rules for structure form selection from the case base and a method for mining fuzzy association rules. In the fuzzy association rule mining, we present a method for fuzzy interval division and fuzzification of quantitative attributes of the real cases. We demonstrate the application of the Apriori algorithm to generate association rules that can be used in building design. This data mining approach provides a new technical support for design efficiency, quality and intelligence. © Springer-Verlag Berlin Heidelberg 2007.