Filtrado espacial, semántico y colaborativo para apoyar decisiones en entornos ubicuos Spatial, semantic and collaborative filtering to support decisions on ubiquitous environments

Recommender Systems have emerged on the Web to customize information received by users. This research provides spatial and semantic components to build Collaborative Recommender Systems. The use of web services derived from the Spatial Data Infrastructure of the Republic of Cuba, which constitutes the main spatial data source accessible via Web in the country, is particularly relevant. The inclusion of concepts from Semantic Web, as ontologies and semantic query services, allow information retrieving used by Recommender Systems engines more effectively. A methodology based on two stages of filters (spatio-semantic and collaborative filters) is described. In the first stage, considered the pre-filtering, spatial and semantic filters are explained. Once the information necessary for the Collaborative Recommender System is obtained, the resulting points of interest and user preferences are processed by the Collaborative Filter. Finally, some elements of a Recommender System implementation for a mobile application are discussed. This work impacts on the need to obtain useful information from distributed, voluminous and heterogeneous data, in a more ubiquitous world, where mobile devices and Web sensors growth exponentially.