A ligação de entidades como uma etapa para a recuperação semântica da informação

This article is about the entity linking as a primordial step for the semantic information retrieval, it’s use analyzing the application of the DBpedia Spotlight software in descriptive texts of postage stamps. The methodology consists in an exploratory research on the objectives, bibliographic and case study on the means. The case study explores the application of software in natural language descriptive texts for entity identification and entity linking to the DBpedia knowledge base. The research reveals the software’s ability to create specialized and interconnected knowledge bases through data originated from Wikipedia. It is concluded that the entity linking is an important step for the semantic retrieval of objects that present in their textual description entities portrayed in the encyclopedia.