The goal of this Phd Thesis is to put forward an evolving semantic model-driven approach, where the underlying conceptual model of the different data sources is preserved. Instead of data conversion, the semantic model-driven approach ensures that for a given application, all the information that can be represented by a given standard will be available to the application. A model-driven approach ensures that even if standards evolve, the application can make use of all the information exposed by any of these data sources. In this PhD Thesis, we will focus on 3D urban data.

Building an evolving semantic model is challenging, since:

1. Choosing an expressive semantic model requires a proper understanding of the different information that can be represented by the different sources and the standards they use.
2. Automated mapping of urban model information to the proposed semantic model has several limitations
3. Automated evolution of the semantic model, with the integration of new data sources as well as new versions of the associated data standards make the problem very difficult.

Both industry and academia are actively working in this area. We want to research this problem in the context of urban data, considering a number of highly evolving data standards for representing diverse n-dimensional urban data. The goal is to be able to ensure data interoperability and the possibility to integrate with other open linked data. This proposal also aims to improve and ensure geospatial data on the web and their querying.