%0 Conference Proceedings
%T Named Entity Recognition and Relation Extraction with Graph Neural Networks in Semi Structured Documents
%A Manuel Carbonell
%A Pau Riba
%A Mauricio Villegas
%A Alicia Fornes
%A Josep Llados
%B 25th International Conference on Pattern Recognition
%D 2020
%F Manuel Carbonell2020
%O DAG; 600.121
%O exported from refbase (http://refbase.cvc.uab.es/show.php?record=3509), last updated on Wed, 27 Jan 2021 13:31:25 +0100
%X The use of administrative documents to communicate and leave record of business information requires of methodsable to automatically extract and understand the content fromsuch documents in a robust and efficient way. In addition,the semi-structured nature of these reports is specially suitedfor the use of graph-based representations which are flexibleenough to adapt to the deformations from the different documenttemplates. Moreover, Graph Neural Networks provide the propermethodology to learn relations among the data elements inthese documents. In this work we study the use of GraphNeural Network architectures to tackle the problem of entityrecognition and relation extraction in semi-structured documents.Our approach achieves state of the art results in the threetasks involved in the process. Additionally, the experimentationwith two datasets of different nature demonstrates the goodgeneralization ability of our approach.
%U http://refbase.cvc.uab.es/files/CRV2020.pdf