By ontological geosciences we mean geosciences that are recognizable by computers and expressed in hierarchical concrete concepts and geoscientific concept models.
Hierarchical geoscientific concepts and concept models have been constructed and discussed for a quite long time in the geoinformatics community. However, apart from geoscientific data integration, the use of conceptualized knowledge is rarely mentioned. This paper will introduce some progress in our ongoing development of multi-dimensional visual concept space for geoscientific data mining.
Traditionally, data cluster analysis techniques are widely used in geosciences but nowadays data mining approaches are beginning to be taken in from the wider information technology research community. One feature that strongly distinguishes data mining from traditional data cluster analysis is the introduction of multi-dimensional space into data cluster analysis. Data in multi-dimensional space can be sliced, diced, rolled up, or drilled down, as well as clustered and classified.
Hierarchical geoscientific concepts naturally represent a multi-dimensional space for data rolling up or drilling down and geoscientific concept models can provide the necessary knowledge to slice and dice data in multi-dimensional space. This paper takes an example of a layered earth concept model to show how to roll up and drill down in hierarchical geoscientific concepts and slice and dice the data using geoscientific concept models.