2013/07/30 by Jason Weston, Weston, Jason, Antoine Bordes +5
Computer Science · Biochemistry, Genetics and Molecular Biology · #Natural Language Processing Techniques #Topic Modeling #Biomedical Text Mining and Ontologies
paper · pdf · doi:10.48550/arxiv.1307.7973
This paper proposes a novel approach for relation extraction from free text\nwhich is trained to jointly use information from the text and from existing\nknowledge. Our model is based on two scoring functions that operate by learning\nlow-dimensional embeddings of words and of entities and relationships from a\nknowledge base. We empirically show on New York Times articles aligned with\nFreebase relations that our approach is able to efficiently use the extra\ninformation provided by a large subset of Freebase data (4M entities, 23k\nrelationships) to improve over existing methods that rely on text features\nalone.\n