Improved lexically constrained decoding for translation and monolingual rewriting
| Autor: | Benjamin Van Durme, Tongfei Chen, Matt Post, Patrick Xia, Huda Khayrallah, Ryan Culkin, J. Edward Hu |
|---|---|
| Předmět: |
Sequence
Phrase Machine translation Computer science business.industry Computer Science::Information Retrieval Computer Science::Computation and Language (Computational Linguistics and Natural Language and Speech Processing) 02 engineering and technology computer.software_genre Constraint (information theory) 030507 speech-language pathology & audiology 03 medical and health sciences 0202 electrical engineering electronic engineering information engineering Question answering 020201 artificial intelligence & image processing Rewriting Artificial intelligence 0305 other medical science business Throughput (business) computer Natural language processing Decoding methods |
| Zdroj: | Scopus-Elsevier NAACL-HLT (1) |
| Popis: | Lexically-constrained sequence decoding allows for explicit positive or negative phrase-based constraints to be placed on target output strings in generation tasks such as machine translation or monolingual text rewriting. We describe vectorized dynamic beam allocation, which extends work in lexically-constrained decoding to work with batching, leading to a five-fold improvement in throughput when working with positive constraints. Faster decoding enables faster exploration of constraint strategies: we illustrate this via data augmentation experiments with a monolingual rewriter applied to the tasks of natural language inference, question answering and machine translation, showing improvements in all three. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|