Top-down string-to-dependency Neural Machine Translation
arXiv:2603.27938v1 Announce Type: new Abstract: Most of modern neural machine translation (NMT) models are based on an encoder-decoder framework with an attention mechanism. While they perform well on standard datasets, they can have trouble in translation of long inputs that are rare or unseen during training. Incorporating target syntax is one approach to dealing with such length-related problems. We propose a novel syntactic decoder that generates a target-language dependency tree in a top-down, left-to-right order. Experiments show that the proposed top-down string-to-tree decoding general — Shuhei Kondo, Katsuhito Sudoh, Yuji Matsumoto
View PDF HTML (experimental)
Abstract:Most of modern neural machine translation (NMT) models are based on an encoder-decoder framework with an attention mechanism. While they perform well on standard datasets, they can have trouble in translation of long inputs that are rare or unseen during training. Incorporating target syntax is one approach to dealing with such length-related problems. We propose a novel syntactic decoder that generates a target-language dependency tree in a top-down, left-to-right order. Experiments show that the proposed top-down string-to-tree decoding generalizes better than conventional sequence-to-sequence decoding in translating long inputs that are not observed in the training data.
Subjects:
Computation and Language (cs.CL)
Cite as: arXiv:2603.27938 [cs.CL]
(or arXiv:2603.27938v1 [cs.CL] for this version)
https://doi.org/10.48550/arXiv.2603.27938
arXiv-issued DOI via DataCite (pending registration)
Submission history
From: Shuhei Kondo [view email] [v1] Mon, 30 Mar 2026 01:21:38 UTC (134 KB)
Sign in to highlight and annotate this article
Conversation starters
Daily AI Digest
Get the top 5 AI stories delivered to your inbox every morning.
More about
researchpaperarxiv
Knowledge Map
Connected Articles — Knowledge Graph
This article is connected to other articles through shared AI topics and tags.
More in Research Papers
Quantum computers might crack today's encryption far sooner than we thought
According to a study by engineers at Caltech and the UC Department of Physics, quantum computers do not need to be nearly as powerful as previously believed to crack the most advanced cryptographic technologies. The research claims that Shor's algorithm could break RSA public-key encryption using quantum computers with just... Read Entire Article
Discussion
Sign in to join the discussion
No comments yet — be the first to share your thoughts!