Classifying Divergences in Cross-lingual AMR Pairs

Wein, Shira; Schneider, Nathan

doi:10.18653/v1/2021.law-1.6

Cited by 7 publications

(7 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recent work classified these AMRs via AMR structural divergence schema (Wein and Schneider, 2021) (mentioned in §2). (Wein and Schneider, 2021) In table 3, we measure the ability of our pipeline system and the and Carpuat, 2020) system to detect semantic divergences at a stricter level, as picked up by the AMR divergence schema.…”

Section: Results On Gold English-french Amr Pairsmentioning

confidence: 99%

See 1 more Smart Citation

Measuring Fine-Grained Semantic Equivalence with Abstract Meaning Representation

Wein¹,

Wang²,

Schneider³

2022

Preprint

View full text Add to dashboard Cite

Identifying semantically equivalent sentences is important for many cross-lingual and monolingual NLP tasks. Current approaches to semantic equivalence take a loose, sentencelevel approach to "equivalence," despite previous evidence that fine-grained differences and implicit content have an effect on human understanding (Roth and Anthonio, 2021) and system performance (Briakou and Carpuat, 2021). In this work, we introduce a novel, more sensitive method of characterizing semantic equivalence that leverages Abstract Meaning Representation graph structures. We develop an approach, which can be used with either gold or automatic AMR annotations, and demonstrate that our solution is in fact finer-grained than existing corpus filtering methods and more accurate at predicting strictly equivalent sentences than existing semantic similarity metrics. We suggest that our finer-grained measure of semantic equivalence could limit the workload in the task of human post-edited machine translation and in human evaluation of sentence similarity.

show abstract

Section: Results On Gold English-french Amr Pairsmentioning

confidence: 99%

“…Recent work has investigated the differences in cross-lingual (English-Spanish) AMR pairs within the framework of translation divergences (Wein and Schneider, 2021). Specifically, this work developed an annotation schema to classify the types and causes of differences between cross-lingual AMR pairs.…”

Section: Background On Semantic Divergencementioning

confidence: 99%

Measuring Fine-Grained Semantic Equivalence with Abstract Meaning Representation

Wein¹,

Wang²,

Schneider³

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…A second limitation of the previous Spanish AMR annotation was the limited amount of change to the English AMR guidelines to incorporate Spanish linguistic phenomena. Recent work has assessed various differences between Spanish and English annotations of the existing Spanish AMR adaptation, classifying the type and cause of the identified differences (Wein and Schneider, 2021).…”

Section: Prior Work Adapting Amr To Spanishmentioning

confidence: 99%

“…to adequately capture semantic phenomena in Spanish, they indicated that accurate representation could be accomplished by adding specific roles and constructions. For example, the English and Spanish AMRs in figure 1, which annotate parallel sentences, have two syntactic divergences due to inherent differences between the languages (Wein and Schneider, 2021).…”

Section: Introductionmentioning

confidence: 99%

Spanish Abstract Meaning Representation: Annotation of a General Corpus

Wein

Donatelli²,

Ethan³

et al. 2022

NEJLT

View full text Add to dashboard Cite

Abstract Abstract Meaning Representation (AMR), originally designed for English, has been adapted to a number of languages to facilitate cross-lingual semantic representation and analysis. We build on previous work and present the first sizable, general annotation project for Spanish AMR. We release a detailed set of annotation guidelines and a corpus of 486 gold-annotated sentences spanning multiple genres from an existing, cross-lingual AMR corpus. Our work constitutes the second largest non-English gold AMR corpus to date. Fine-tuning an AMR to-Spanish generation model with our annotations results in a BERTScore improvement of 8.8%, demonstrating initial utility of our work.

show abstract

“…However, it has to face two limitations (at least): relying on the translation quality, which is not the same for all languages, and the possible domain-level divergence between the translated and the gold corpora, which can prevent leveraging all its potential. Concerning AMR, we also can suffer from some structural divergences in the translated AMR graphs (Sobrevilla Cabezudo et al, 2019;Wein and Schneider, 2021) as, originally, AMR is highly biased to English (Banarescu et al, 2013).…”

mentioning

confidence: 99%

Comparison of Cross-lingual Strategies for AMR-to-Brazilian Portuguese Generation

Cabezudo

Anchiêta

Pardo

2022

Preprint

View full text Add to dashboard Cite

This work presents a study of different strategies for tackling low-resource AMR-to-text generation for Brazilian Portuguese on three approaches. In particular, we explore the use of English AMR as an interlingua and transfer learning (TL). The results suggest that using AMR as an interlingua can be a strong baseline. However, we need to consider the bilingual dictionary used in the concept translation, as it can harm text generation. On the other hand, TL seems to be a promising strategy for generating accurate outputs, but its contribution is not significant. Finally, we present a transformer-based model with the best performance, surpassing all baselines.

show abstract

Classifying Divergences in Cross-lingual AMR Pairs

Cited by 7 publications

References 18 publications

Measuring Fine-Grained Semantic Equivalence with Abstract Meaning Representation

Measuring Fine-Grained Semantic Equivalence with Abstract Meaning Representation

Spanish Abstract Meaning Representation: Annotation of a General Corpus

Comparison of Cross-lingual Strategies for AMR-to-Brazilian Portuguese Generation

Contact Info

Product

Resources

About