Geospatial Semantic Similarity

In recent times, the development of geographic information systems has reached a high degree of maturity due to the perfecting of the underlying technologies. In this sense, it is worth mentioning how artificial intelligence and machine learning techniques have been able to solve some of the problems related to semantic interoperability in the context of such information systems.

If in the past, it was necessary to resort to external dictionaries to discern the degree of semantic similarity of terms such as, for example, hotel and tourist apartment. Now, new techniques based on the use of neural networks and fuzzy logics allow to obtain quite precise results in this context.

The framework of our work is focused on the processing of landmarks and points of interest, to carry out tasks such as content recommendation. We have been working on the intelligent aggregation of semantic similarity techniques that have been giving very good results lately. We hope that in the next few years, these techniques can even be improved, so that the geographic information systems can become better and reach the general public.