Semantic Web Based Named Entity Linking for Digital Humanities and Heritage Texts

International audienceThis paper proposes a graph based methodology for automaticallydisambiguating authors’ mentions in a corpus of French literarycriticism. Candidate referents are identified and evaluated using a graphbased named entity linking algorithm, which exploits a knowledge-basebuilt out of two different resources (DBpedia and the BnF linked data).The algorithm expands previous ones applied for word sense disambiguationand entity linking, with good results. Its novelty resides in the factthat it successfully combines a generic knowledge base such as DBpediawith a domain specific one, thus enabling the efficient annotation ofminor authors. This will help specialists to follow mentions of the sameauthor in different works of literary criticism, and thus to investigatetheir literary appreciation over time