BSC Participation in the WMT Translation of Biomedical Abstracts

Abstract

This paper describes the machine translation systems developed by the Barcelona Supercomputing (BSC) team for the biomedical translation shared task of WMT19. Our system is based on Neural Machine Translation unsing the OpenNMT-py toolkit and Transformer architecture. We participated in four translation directions for the English/Spanish and English/Portuguese language pairs. To create our training data, we concatenated several parallel corpora, both from in-domain and out-of-domain sources, as well as terminological resources from UMLS.

Automatisiertes klinisches Codieren

“…To ease the comparison with such systems, provide support to participants working previously in English and to explore the use of machine-translated corpora, we generated a machine translated version of the CodiEsp corpus (CodiEsp MT corpus). The used machine translation system was adapted to the language characteristics of the medical domain [42].…”

“…This enabled us to build a mapping from DeCS terms to ICD-10 codes with this mapping chain PubMed machine-translation 8 . A large collection of PubMed abstracts was automatically translated into Spanish using the same translation engine employed to translate the CodiEsp corpus [42]. PubMed abstracts are manually indexed with MeSH and easily mapped to ICD-10 terms using the same mapping chain employed for CodiEsp-abstracts.…”

Findings of the WMT 2019 Biomedical Translation Shared Task: Evaluation for MEDLINE Abstracts and Biomedical Terminologies

“…BCS (Soares and Krallinger, 2019). The team's systems were also based on the Transformer-big architecture, which were trained using the OpenNMT-py toolkit.…”

SciBabel: a system for crowd-sourced validation of automatic translations of scientific texts

“…Translation quality increased by around 51% (or 16 percentage points) from 2016 to 2019 for English to Spanish. In Table 1 [11][12][13][14][15][16], we show the MT performance for some language pairs for biomedical texts with dates ranging from 2013 to 2019 for selected language pairs. Note that translation quality is measured automatically using the BLEU score, a common MT metric that relies on the overlapping portions of the generated translations and the manually translated text [17].…”