Today, the amount and variety of spatial data have increased dramatically. In addition, the web has made it easier to disseminate and share this kind of data. Therefore, spatial data integration and interoperability have gained more importance. Spatial data are collected from different sources and often heterogeneous in terms of the levels of detail and the points of view. To able to meet the demands of different spatial applications, multi-source and heterogeneous spatial datasets need to be integrated as well as the consistency of these datasets needs to be maintained. In this context, multi-representation spatial database (MRSDB) paradigm has been suggested by researchers. However, the heterogeneity constitutes a significant barrier in this respect and hence the implementations have so far been remained within a rather narrow scope. In this article, it is mainly discussed about the possible contributions of basic methods and technologies of spatial semantics such as ontologies, semantic web and linked data to the data integration for creating a MRSBD. Some examples are also given to illustrate the concept.