Around the world, land cover changes occur due to natural and anthropogenic factors. In many cases, the consequences of anthropogenic interventions are unexpected. In order for scientists and policy makers to identify land cover change processes of interest, it is necessary suitable tools for early and efficient analysis of land cover data. In our research, we present a data model that makes use of semantic web technologies to manage a hierarchical structure of land cover types. Using this approach, it is possible to manage the land cover information at different levels of abstraction. In our research, we use a Change Vector Analysis approach to represent the land cover change. The number and semantics of the axes in the resulting multidimensional space depend on the level of abstraction decided by the researcher and can be modified on the fly. Using this approach, it is possible to detect processes allowing the researcher to have quantitative change measurements. In this work, we present our research using data from the CORINE land cover program, corresponding to the years 1990, 2000 and 2006. As our study area, we have selected ten countries in Eastern Europe, that became members of the EU between 2005 and 2007. Our results suggest that our method is efficient, reliable and scalable.