This study presents the creation of semantic correspondence rules for an attribute of interest and the framework for the analysis of thematic data semantic similarity between corresponding building entities in different databases. The analysis was based on the Polish Database of Topographic Objects, the official National Mapping Agency dataset, and Volunteered Geographic Information available from the OpenStreetMap (OSM) project. The work required proposing a user oriented approach to tackle several specific issues, namely: variable data conceptualizations and content organizations manners, and unambiguous one-to-one correspondence lack between two classification schemas, patchy OSM completeness, and language medium -- British English vs. Polish.