We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CL

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computation and Language

Title: Measuring Alignment Bias in Neural Seq2Seq Semantic Parsers

Abstract: Prior to deep learning the semantic parsing community has been interested in understanding and modeling the range of possible word alignments between natural language sentences and their corresponding meaning representations. Sequence-to-sequence models changed the research landscape suggesting that we no longer need to worry about alignments since they can be learned automatically by means of an attention mechanism. More recently, researchers have started to question such premise. In this work we investigate whether seq2seq models can handle both simple and complex alignments. To answer this question we augment the popular Geo semantic parsing dataset with alignment annotations and create Geo-Aligned. We then study the performance of standard seq2seq models on the examples that can be aligned monotonically versus examples that require more complex alignments. Our empirical study shows that performance is significantly better over monotonic alignments.
Comments: 5 pages, 3 figures, accepted at *SEM 2022: The 11th Joint Conference on Lexical and Computational Semantics
Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
Cite as: arXiv:2205.08288 [cs.CL]
  (or arXiv:2205.08288v1 [cs.CL] for this version)

Submission history

From: Davide Locatelli [view email]
[v1] Tue, 17 May 2022 12:35:52 GMT (7245kb,D)

Link back to: arXiv, form interface, contact.