Background knowledge and PMML
Tomáš Kliegr, Stanislav Vojíř, Jan Rauch · 2011
This paper introduces an attempt for a holistic approach to integrating background knowledge with PMML in the area of association rule mining. An XML format for capturing background knowledge inspired by PMML that scopes pre-processing and background knowledge association rules is presented. We argue that background knowledge should be kept separated from a specific dataset to foster its reusability and propose an XML format that addresses the problem of interlinking with background knowledge. The discussion is motivated by our experience from development of domain expert oriented data mining support platform.