A Theory of Attributed Equivalence in Databases with Application to Schema Integration
IEEE Transactions on Software Engineering
Neural Networks: A Comprehensive Foundation
Neural Networks: A Comprehensive Foundation
The Representation Race - Preprocessing for Handling Time Phenomena
ECML '00 Proceedings of the 11th European Conference on Machine Learning
A Comparison of Ranking Methods for Classification Algorithm Selection
ECML '00 Proceedings of the 11th European Conference on Machine Learning
| Hi-index | 0.00 |
Although some research has been dedicated to the development of Knowledge Discovery in Databases (KDD) assistance mechanisms, little effort has been directed to the deployment of tools that assist humans during the KDD task definition stage. In order to satisfy this need for a KDD task definition assistance device, the present work proposes three different approaches: a) the first one is called theoretical approach and is based on concepts from the Theory of Attribute Equivalence in Databases [3] and from Topological Spaces [4]; b) the second employs Artificial Neural Networks [7] to learn mappings between heterogeneous patterns and is called experimental approach; c) the third one combines the abovementioned approaches to implement what is called hybrid approach. These approaches, their models and implementations are described in detail. Experiments with real KDD applications, comparisons and conclusions are reported.