Abstract:

The purpose of the study is to develop a framework that integrates knowledge management (KM) and decision support systems (DSS) by using knowledge discovery techniques (KDT). The framework applies KDT for various types of knowledge conversion and also generates specific models by utilizing previously defined models as much as possible. Knowledge externalization is achieved by extracting decision rules from previously trained artificial neural network models that are stored in model base. Explicit knowledge in the form of decision rules are combined by a grid based density clustering algorithm, called CLIQUE, based on predefined similarity measures. The rest of the knowledge conversion types, internalization and socialization, are performed by case based reasoning (CBR) paradigm. CBR also enables to utilize specific models for each of the problem automatically. In order to prove the applicability of the proposed framework an experimental study is designed to forecast the change in the Turkish financial and macroeconomic time series.