Linking Four Heterogeneous Language Resources as Linked Data
Benjamin Siemoneit, John Philip McCrae, Philipp Cimiano · 2015
The interest in publishing language resources as linked data is increasing, as clearly corroborated by the recent growth of the Linguistic Linked Data cloud.However, the actual value of data published as linked data is the fact that it is linked across datasets, supporting integration and discovery of data.As the manual creation of links between datasets is costly and therefore does not scale well, automatic linking approaches are of great importance to increase the quality and degree of linking of the Linguistic Linked Data cloud.In this paper we examine an automatic approach to link four different datasets to each other: two terminologies, the European Migration Network (EMN) glossary as well as the Interactive Terminology for Europe (IATE), BabelNet, and the Manually Annotated Subcorpus (MASC) of the American National Corpus.We describe our methodology, present some results on the quality of the links and summarize our experiences with this small linking exercise We will make sure that the resources are added to the linguistic linked data cloud.