Data Mining and Machine Oriented Modeling: A Granular Computing Approach
Applied Intelligence
Frameworks for Mining Binary Relations in Data
RSCTC '98 Proceedings of the First International Conference on Rough Sets and Current Trends in Computing
Granulation and nearest neighborhoods: rough set approach
Granular computing
Granular computing: structures, representations, and applications
RSFDGrC'03 Proceedings of the 9th international conference on Rough sets, fuzzy sets, data mining, and granular computing
Hi-index | 0.00 |
Topological data models. One of the most important goals of data modeling is to capture the information requirements of an application in terms of structures, constrains, and operations that naturally reflect the real world situation. To date Codd's relational data modeling has been one of the most successful representations of real world problems. As requirements are extended from "(exact) retrievals" or "(exact) queries" to "approximate retrievals" or "fuzzy queries". There are needs to generalize relational data models. To capture certain semantics information, various authors introduce new structures into relational data models: Ginsburg and Hull introduced partial ordering, while Motro introduced distance. Analysis of some examples of Motro's approach reveals that the exact numeric values of distances are not important. The essence lies in the qualitative massage carried by the numeric values. In this paper, the notion of topology is used to capture the qualitative messages directly. The model is called topological data model. In terms of category theory, Codd's relational data models are representations of real world problems in the category of sets, Ginsburg and Hull's models are in the category of partial ordered sets, Motro's models are in the category of metric spaces, while present work are in the category of topological spaces.