This paper presents our solution for the BUCC 2018 Shared Task on parallel sentence extraction from comparable corpora. Our system identifies parallel sentence pairs in French-English corpora by following a hybrid approach pairing multilingual sentence-level embeddings, neural machine translation, and supervised classification. Our system consists of a two-step process. In the first step, to reduce the size and the noise of the candidate sentence pairs, we filter the target translation candidates using the continuous vector representation of each source-target sentence pair learned using a bilingual distributed representation model. Then we select the best translation using a neural machine translation system or a binary classification model. We achieve an F1-score of up to 75.2 and 76.0 on the BUCC18 train and test sets respectively.
@InProceedings{BOUAMOR18.8, author = {Houda Bouamor and Hassan Sajjad}, title = {H2@BUCC18: Parallel Sentence Extraction from Comparable Corpora Using Multilingual Sentence Embeddings}, booktitle = {Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018)}, year = {2018}, month = {may}, date = {7-12}, location = {Miyazaki, Japan}, editor = {Reinhard Rapp and Pierre Zweigenbaum and Serge Sharoff}, publisher = {European Language Resources Association (ELRA)}, address = {Paris, France}, isbn = {979-10-95546-07-8}, language = {english} }
Summary of the paper