Hitachi at MRP 2020: Text-to-Graph-Notation Transducer

Abstract

This paper presents our proposed parser for the shared task on Meaning Representation Parsing (MRP 2020) at CoNLL, where participant systems were required to parse five types of graphs in different languages. We propose to unify these tasks as a text-to-graph-notation transduction in which we convert an input text into a graph notation. To this end, we designed a novel Plain Graph Notation (PGN) that handles various graphs universally. Then, our parser predicts a PGN-based sequence by leveraging Transformers and biaffine attentions. Notably, our parser can handle any PGN-formatted graphs with fewer frameworkspecific modifications. As a result, ensemble versions of the parser tied for 1st place in both cross-framework and cross-lingual tracks. * Contributed equally. Ozaki mainly developed PGN parser. Morio mainly developed neural models.

SGL: Speaking the Graph Languages of Semantic Parsing via Multilingual Translation

“…In this sense, we continue the recent cross-framework trend formally started by the shared task of Oepen et al (2019), exploring the possibility of using translation-based approaches for framework-independent parsing, as opposed to the transition-based parsers proposed in that seminal work. Our findings are in line with the recent results reported by Oepen et al (2020) and, in particular, by Ozaki et al (2020) (d) UCCA Linearization Figure 1: AMR and UCCA graphs, along with their linearizations, for the sentence "After graduation, John moved to Paris". To ease readability, linearizations are shown with newlines and indentation; however, when fed to the neural model, they are in a single-line single-space format.…”

ÚFAL at MRP 2020: Permutation-invariant Semantic Parsing in PERIN

“…According to the official whole-percent-only all F1 score, our competition submission reached tied first place in both the cross-lingual and the crossframework track, with its performance virtually (Oepen et al, 2019). identical to the system by Hitachi (Ozaki et al, 2020). Our bugfixed submission reached the first rank in both tracks, improving the cross-lingual score by nearly one percent point.…”

MRP 2020: The Second Shared Task on Cross-Framework and Cross-Lingual Meaning Representation Parsing

“…When parsing cross-framework meaning representations for English, the system is trained with a BERT-large-cased pretrained encoder, and when parsing cross-lingual meaning representations, it is trained with multilingual BERT. (Chen et al, 2019), HIT-SCIR (Che et al, 2019), and Saarland (Donatelli et al, 2019), respectively; in MRP 2020 the Hitachi system (Ozaki et al, 2020) was at the top for all three frameworks, sharing the UCCA first rank withÚFAL (Samuel and Straka, 2020).…”