Edinburgh Associative Thesaurus as RDF and DBpedia Mapping.

Associations, which are one of the key ingredients of human intelligence and thinking, are not easily accessible to the Semantic Web community. High quality RDF datasets of this kind are missing. In this paper we generate such a dataset by transforming 788 K free-text associations of the Edinburgh Associative Thesaurus (EAT) into RDF. Furthermore, we provide a verified mapping of strong textual associations from EAT to DBpedia Entities with the help of a semi-automatic mapping approach. Both generated datasets are made publicly available and can be used as a benchmark for cross-type link prediction and pattern learning.