Interoperable multimedia annotation and retrieval for the tourism sector

Abstract

The Atlas Metadata System (AMS) employs semantic web annotation techniques in order to create an interoperable information annotation and retrieval platform for the tourism sector. AMS adopts state-of-the-art metadata vocabularies, annotation techniques and semantic web technologies. Interoperability is achieved by reusing several vocabularies and ontologies, including Dublin Core, PROV-O, FOAF, Geonames, Creative commons, SKOS, and CiTO, each of which provides with orthogonal views for annotating different aspects of digital assets. Our system invests a great deal in managing geospatial and temporal metadata, as they are extremely relevant for tourism-related applications. AMS has been implemented as a graph database using Neo4j, and is demonstrated with a dataset of more than 160000 images downloaded from Flickr. The system provides with online recommendations, via queries that exploit social networks, spatiotemporal references, and user rankings. AMS is offered via service-oriented endpoints using public vocabularies to ensure reusability.