XL-AMR: Enabling Cross-Lingual AMR Parsing with Transfer Learning Techniques

Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, pages 2487–2500, November 16–20, 2020. c⃝2020 Association for Computational Linguistics 2487 XL-AMR: Enabling Cross-Lingual AMR Parsing with Transfer Learning Techniques Rexhina Blloshmi Rocco Tripodi Roberto Navigli Sapienza NLP Group Department of Computer Science, Sapienza University of Rome {blloshmi,tripodi,navigli}@di.uniroma1.it Abstract Abstract Meaning Representation (AMR) is a popular formalism of natural language that rep- resents the meaning of a sentence as a seman- tic graph. It is agnostic about how to derive meanings from strings and for this reason it lends itself well to the encoding of seman- tics across languages. However, cross-lingual AMR parsing is a hard task, because training data are scarce in languages other than English and the existing English AMR parsers are not directly suited to being used in a cross-lingual setting. In this work we tackle these two problems so as to enable cross-lingual AMR parsing: we explore different transfer learning techniques for producing automatic AMR anno- tations across languages and develop a cross- lingual AMR parser, XL-AMR. This can be trained on the produced data and does not rely on AMR aligners or source-copy mechanisms as is commonly the case in English AMR pars- ing. The results of XL-AMR significantly sur- pass those previously reported in Chinese, Ger- man, Italian and Spanish. Finally we provide a qualitative analysis which sheds light on the suitability of AMR across languages. We re- lease XL-AMR at github.com/SapienzaNLP/xl- amr. 1 Introduction Abstract Meaning Representation (AMR) is a pop- ular formalism for natural language (Banarescu et al., 2013). It represents sentences as rooted, directed and acyclic graphs in which nodes are concepts and edges are semantic relations among them. AMR unifies, in a single structure, a rich set of information coming from different tasks, such as Named Entity Recognition (NER), Semantic Role Labeling (SRL), Word Sense Disambiguation (WSD) and coreference resolution. Such repre- sentations are actively integrated in several Natu- ral Language Processing (NLP) applications, inter alia, information extraction (Rao et al., 2017), text summarization (Hardy and Vlachos, 2018; Liao et al., 2018), paraphrase detection (Issa et al., 2018), spoken language understanding (Damonte et al., 2019), machine translation (Song et al., 2019b) and human-robot interaction (Bonial et al., 2020). It is therefore desirable to extend AMR semantic rep- resentations across languages along the lines of cross-lingual representations for grammatical anno- tation (de Marneffe et al., 2014), concepts (Conia and Navigli, 2020) and semantic roles (Akbik et al., 2015; Di Fabio et al., 2019). Furthermore, it could be especially useful to integrate cross-lingual se- mantic structures in multilingual applications of natural language understanding. A peculiar feature of the AMR formalism is that it aims at abstracting away from word forms. AMR graphs are unanchored, i.e., the linkage be- tween tokens in a sentence and nodes in the corre- sponding graph is not explicitly annotated. Hence, the feature of being agnostic about how to derive meanings from strings makes AMR particularly suitable for representing semantics cross-lingually. However, AMR was initially designed …

Content truncated to 500 words