Recurrent models and lower bounds for projective syntactic decoding

Publikation: Konference artikel i Proceeding eller bog/rapport kapitelKonferencebidrag i proceedingsForskningpeer review


The current state-of-the-art in neural graphbased parsing uses only approximate decoding at the training phase. In this paper aim to understand this result better. We show how recurrent models can carry out projective maximum spanning tree decoding. This result holds for both current state-of-the-art models for shiftreduce
and graph-based parsers, projective or not. We also provide the first proof on the
lower bounds of projective maximum spanning tree, DAG, and digraph decoding.
TitelProceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies
Antal sider10
Vol/bindVolume 1 (Long and Short Papers)
ForlagAssociation for Computational Linguistics
StatusUdgivet - 2019


Dyk ned i forskningsemnerne om 'Recurrent models and lower bounds for projective syntactic decoding'. Sammen danner de et unikt fingeraftryk.