Research map / Label-efficient and robust learning / Recognition and long-tail
Improved Neural Machine Translation with SMT Features
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add word translation table and word reward features (ablation) | GroundHog NMT · NIST MT08 | BLEU: 30.61 -> 31.57 | Helped |
| add 5-gram language model feature (ablation) | GroundHog+TM+WR · NIST MT08 | BLEU: 31.57 -> 32.94 | Helped |
| use translation table to translate OOV words (ablation) | GroundHog+TM+WR+LM · NIST MT08 | BLEU: 32.57 -> 32.94 | Helped |
| integrate SMT features in log-linear NMT | GroundHog · NIST MT08 | OOV percentage: 4.6% -> 0.9% | Helped |
| increase beam size beyond 3 | GroundHog · NIST MT08 | BLEU | No clear effect |
| increase beam size | GroundHog+SMT features · NIST MT08 | BLEU | Helped |
Closest papers by meaning
- Neural Machine Translation Advised by Statistical Machine Translation (AAAI 2017 · similarity 0.97)
- Unsupervised Neural Machine Translation with SMT as Posterior Regularization (AAAI 2019 · similarity 0.91)
- Neural Machine Translation with Reconstruction (AAAI 2017 · similarity 0.90)
- Joint Training for Neural Machine Translation Models with Monolingual Data (AAAI 2018 · similarity 0.90)
- Neural Machine Translation with Joint Representation (AAAI 2020 · similarity 0.89)
- Prompting Neural Machine Translation with Translation Memories (AAAI 2023 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.