The Association Analysis for Construction Laws and Regulations Data
Bianping Su, Hui Zhi, Yi-ping Wang · 2009
Association analysis is especially important to users. It is becoming very important to provide association recommendation by the association analysis of construction laws and regulations data for the users. Extracted features from each document, the cumulative frequency of 85% of the requirement that has made the feature bearing the most message of the text, summary of all text feature form the feature set of laws and regulations of construction data and make it impossible for association analysis of outlier text features. In gray correlation analysis, the system features is corresponding to the features of laws and regulations of construction data, the corresponding 'sequence item' is corresponding to 'text' that the sub-vector of feature vector. Furthermore, according to the features of construction laws and regulations and referring to the modeling thinking of text vector space model, the vector space model of the data characteristics is established. From analyzing the definition of the frequency item sets, the method to obtain the data characteristic frequent item sets laws and regulations of construction is expounded. From the angle of data characteristic association the generalized gray absolute association degree is improved. Then the association rules of data characteristic are educed and the recall rate of construction laws and regulations data is raised. The effectiveness of this association analysis method is verified on the experiments of construction laws and regulations date sets of Shannxi province.