Contribución a la representación del conocimiento y la explotación de datos sobre contaminación lumínica

This work investigates whether it is possible to gather, organize, analyze, and share in-formation on light pollution by following standard Linked Data and Semantic Web practices.We propose and implement an RDF knowledge graph that integrates night-sky bright-ness measurements and auxiliary metadata by reusing established vocabularies—SOSA/SSN,QUDT, TIME, GeoSPARQL, and SKOS—and a scheme of persistent URIs. The method-ology combines modeling and serialization in Turtle and RDF/XML, integration via repro-ducible scripts in Python and R, loading and querying in GraphDB, and validation throughSPARQL queries and syntactic checks.The graph is enriched with contextual information from diverse sources—urban classi-fication, astronomical seasons, day of the week, events, and special periods such as Christ-mas—which enables expressive querying without bespoke Extract–Transform–Load (ETL)pipelines. We design a SPARQL cookbook for temporal and spatial analyses that definesnightly profiles, territorial comparisons, and event impacts. We also propose an RDF-basedstatistics schema covering basic descriptors (boxplot, mean, mode, standard deviation) withprovenance. We report quality and coverage metrics and document modeling decisionsthat facilitate extension and interoperability (FAIR).The results demonstrate the feasibility of the semantic approach to integrate hetero-geneous sources and enable reproducible analyses.