Data Bases, the Base for Data Mining

Abstract

Data collections provide a basis for solving numerous problems by data mining approach. The advantages of data mining consists in the retrieving of a new knowledge from existing information. The comprehensiveness of the data collection, the structure and quality of the data, and the selection of relevant data sets are extremely important to get correct results. In the crystallographic field, scientists will find several databases dealing with crystal structures of inorganic and organic compounds, or proteins. Usually databases have detailed data evaluation mechanisms integrated in their database production process and offer comprehensive and reliable data sets. The CIF standard enables the scientists to exchange the data. As an example, the Inorganic Crystal Structure Database (ICSD), a source of information for crystallographers, mineralogists, physicists, and chemists will be presented here. The ICSD contains about 120,000 entries (March 2009) of fully determined crystal structures. This chapter gives a detailed description of data collection, the contents of the data fields, data evaluation, and finally search the functionality of the ICSD database.