Models and data assimilation: IG_PhysicalProcess Data Mining:

© ISO 2001 All rights reserved 38 Figure 16 - IG_KnowledgeBase

8.4.3.2 Models and data assimilation: IG_PhysicalProcess

Geographic imagery models are computer based mathematical models that realistically simulate spatially distribute time dependent environmental processes in nature. Simulations are of physical phenomena. Output of model is an IG_Image or other type of geographic information. Data assimilation, a type of modeling is the melding of observations with model simulations to provide accurate estimation of the state of the atmosphere, oceans, and land-surface, etc. The term model reflects that any natural phenomena can only be described to a certain degree of accuracy and correctness. It is important to seek the simplest and most general description that still describes the observations with minimum deviations. It is the power and beauty of the basic laws of physics that even complex phenomena can be understood and quantitatively be described on the base of a few simple and general principles. Editors note: Investigate relevance of ISOIEC JTC 1SC24 projects to this clause.

8.4.3.3 Data Mining:

IG_InferenceRules Data mining is the process of discovering hidden, previously unknown and usable correlations in data. The data is analyzed without the necessity of any hypothesis expected result. Data mining delivers knowledge that can be used for a better understanding of the data. ISOIEC 13249-6:2002 During the last decades, imaging satellite sensors have acquired huge quantities of data. Optical, synthetic aperture radar SAR, and other sensors have delivered several millions of scenes that have been systematically collected, processed, and stored. The state-of-the-art systems for accessing remote sensing data and images, in particular, allow only queries by geographical coordinates, time of acquisition, and sensor type. This information is often less relevant than the content of the scene, e.g., structures, patterns, objects, or scattering properties. Thus, only few of the acquired images can actually be used. In the future, the access to image archive will become more difficult due to the enormous data quantity acquired by a new generation of high-resolution satellite sensors. As a consequence, new © ISO 2001 All rights reserved 39 technologies are needed to easily and selectively access the information content of image archives and finally to increase the actual exploitation of satellite observations. [Datcu, et. al.] Data mining, also known as knowledge discovery from databases, is the higher-level process of obtaining information through distilling information into knowledge ideas and beliefs about the mini-world through interpretation of information and integration with existing knowledge. Data mining is concerned with investigators formulating new predictions and hypotheses from data as opposed to testing deductions from theories through a sub-process of induction from a scientific database. Data mining uses an IG_KnoweledgeBase as a repository that integrates IG_Image from one or more sources. In contrast to transactional database design, good IG_KnoweledgeBase design maximizes the efficiency of analytical data processing or data examination for decision making. In addition to data mining, a IG_KnoweledgeBase often supports online analytical processing OLAP tools. OLAP tools provide multidimensional summary views of the IG_KnoweledgeBase, e.g., roll-up increasing the level of aggregation, drill-down decreasing the level of aggregation, slice and dice selection and projection and pivot re-orientation of the multidimensional data view. As a specialization of data mining for broader information, geospatial specifics are utilized. Examples include: geographic measurement frameworks geometry and topology; spatial dependency and heterogeneity; complexity of spatio-temporal objects and rules; and diverse datatypes for geographic imagery. Editors note: an item for IT Roadmap “Languages to describe data mining patterns; Describe patterns to be found and those found.” Relate to ISO 19109 Feature Cataloguing?

8.4.3.4 Feature fusion