[4Rin1] Interactive Session 2

[4Rin1-12] Predicate argument structure analysis with Pointer Networks

〇Keigo Takahashi¹, Hikaru Omori¹, Mamoru Komachi¹ (1. Tokyo Metropolitan University)

Keywords:Natural language processing, Predicate argument structure analysis, Pointer Networks

Recently, neural network models are used in intra-sentential predicate argument structure analysis (PASA).
However, the sequence labeling approach may assign the same case for multiple arguments. Thus, we propose a
novel PASA method by using the Pointer Networks and alleviate the problem of multiple case assignments.