Hierarchical Document Encoder for Parallel Corpus Mining
arXiv:1906.08401
Abstract
We explore using multilingual document embeddings for nearest neighbor mining of parallel data. Three document-level representations are investigated: (i) document embeddings generated by simply averaging multilingual sentence embeddings; (ii) a neural bag-of-words (BoW) document encoding model; (iii) a hierarchical multilingual document encoder (HiDE) that builds on our sentence-level model. The results show document embeddings derived from sentence-level averaging are surprisingly effective for clean datasets, but suggest models trained hierarchically at the document-level are more effective on noisy data. Analysis experiments demonstrate our hierarchical models are very robust to variations in the underlying sentence embedding quality. Using document embeddings trained with HiDE achieves state-of-the-art performance on United Nations (UN) parallel document mining, 94.9% P@1 for en-fr and 97.3% P@1 for en-es.
accepted by WMT2019
References in corpus (5)
- Distributed Representations of Sentences and Documents
- Achieving Human Parity on Automatic Chinese to English News Translation
- Document Embedding with Paragraph Vectors
- A Deep Neural Network Approach To Parallel Sentence Extraction
- Representing Documents and Queries as Sets of Word Embedded Vectors for Information Retrieval