Analysing Typology of European Energy Structure by Using Clustering Algorithms : Proceedings of the 12th International Conference on Informatics in Economy Education, Research & Business Technologies, 25-28 April 2013

The paper proposes to identify the energetic profile for the European countries by using of the nine indicators concerning the energetic sector. The European countries were grouped using the clustering and the feature extraction techniques. The analysis results were visualized on maps in a geographic information system (GIS). The used algorithms for grouping are bisecting KMeans and Non Negative Matrix Factorization. These are provided by the Java Data Mining API framework in the Oracle implementation. The countries groups are viewed on maps that are built in ArcGIS by ESRI. The proposed IT solution in this paper mixed the various solutions in a unitary manner and contained of the specialized elements for programming Java applications. It's about programming data mining applications using Oracle Data Mining Java API and the programming elements through ArcObjects SDK for Java. This two frameworks (Data Mining Java API and ArcObjects SDK) are used because of the Java background and on the other hand, Oracle offers the necessary support for the spatial database and the Data Mining engine.