Extraction of temporal information of the DBpedia: Integration proposal in a semi-structured corpus