Published: 1996 August
Buchtitel: Proceedings of the 2nd International Conference on Knowledge Discovery in Databases (KDD'96), Portland, Oregon, August 2 - 4, 1996
Performing the complex task of Knowledge Discovery in Databases (KDD) requires a break-down of the task-complexity to enable the possibility of performing the KDD-task. Since even more techniques will appear in the future that can solve a variety of KDD-problems, a domain expert that wants to analyse his domain should have the means to work with tools that integrate several of these techniques as well as the techniques themselves. In this paper a framework is proposed for a strategy component that is to be used for a KDD-system that can guide users in breaking down the complexity of a typical KDD-task and supports him in selecting and using several ML-techniques. The goals of such a guidance component are reuse of (predefined) taskcomponents in order to decrease development time and to simplify the process of decomposing a KDD-task, task-oriented planning in order to break down complexity of a typical KDD-task and supporting post-processing (evaluation) of KDD-processes.