R.R. Goldschmidt and E.P. Lopes Passos (Brazil)
Knowledge discovery in databases; data dining;intelligent knowledge discovery assistance; planning;knowledge representation.
Current state of the art in KDD – Knowledge Discovery in Databases – shows that many algorithms are available for use in KDD steps. Each algorithm has its own characteristics and purposes. Considering the huge diversity of KDD algorithms, selecting an adequate sequence of algorithms to be used in a KDD problem becomes a non-trivial decision. In order to help men in making such decision, the present work aims at proposing a computational model that assists humans in the KDD planning process. The proposed approach uses configurable KDD knowledge, represented in KDD ontologies and in heuristic functions.
Important Links:
Go Back
