Skip to content
conftrace_
2019 ACL ACL 2019

Noisy Parallel Corpus Filtering through Projected Word Embeddings

AbstractWe present a very simple method for parallel text cleaning of low-resource languages, based on projection of word embeddings trained on large monolingual corpora in high-resource languages. In spite of its simplicity, we approach the strong baseline system in the downstream machine translation evaluation.

🌉 Interdisciplinary Bridge - Artificial Intelligence and Natural Language Processing
🐣 Hot Topic Early Bird - low-resource language

More from ACL 2019

All ACL 2019 papers →