Adaptation of machine translation for multilingual information retrieval in the medical domain

Abstract

Objective. We investigate machine translation (MT) of user search queries in the context of cross-lingual information retrieval (IR) in the medical domain. The main focus is on techniques to adapt MT to increase translation quality; however, we also explore MT adaptation to improve effectiveness of cross-lingual IR.Methods and Data. Our MT system is Moses, a state-of-the-art phrase-based statistical machine translation system. The IR system is based on the BM25 retrieval model implemented in the Lucene search engine. The MT techniques employed in this work include in-domain training and tuning, intelligent training data selection, optimization of phrase table configuration, compound splitting, and exploiting synonyms as translation variants. The IR methods include morphological normalization and using multiple translation variants for query expansion. The experiments are performed and thoroughly evaluated on three language pairs: Czech-English, German-English, and French-English. MT quality is evaluated on data sets created within the Khresmoi project and IR effectiveness is tested on the CLEF eHealth 2013 data sets.Results. The search query translation results achieved in our experiments are outstanding -our systems outperform not only our strong baselines, but also Google Translate and Microsoft Bing Translator in direct comparison carried out on all the language pairs. The baseline BLEU scores increased from 26.59 to 41.45 for Czech-English, from 23.03 to 40.82 for German-English, and from 32.67 to 40.82 for French-English. This is a 55% improvement on average. In terms of the IR performance on this particular test collection, a significant improvement over the baseline is achieved only for French-English. For Czech-English and German-English, the increased MT quality does not lead to better IR results.Conclusions. Most of the MT techniques employed in our experiments improve MT of medical search queries. Especially the intelligent training data selection proves to be very successful for domain adaptation of MT. Certain improvements are also obtained from German compound splitting on the source language side. Translation quality, however, does not appear to correlate with the IR performance -better translation does not necessarily yield better retrieval. We discuss in detail the contribution of the individual techniques and state-of-the-art features and provide future research directions.

Findings of the 2014 Workshop on Statistical Machine Translation

“…We used the CLEF 2013 eHealth Task 3 test collection containing about 1 million web pages (in English), 50 test queries (originally in English and translated to Czech, German, and French), and their relevance assessments. Some of the participants of the WMT Medical Task (three teams with five submissions in total) submitted translations of the queries (from Czech, German, and French) into English and these translations were used to query the CLEF 2013 eHealth Task 3 test collection using a state-of-the-art system based on a BM25 model, described in Pecina et al (2014). Originally, we asked for 10 best translations for each query, but only the best one were used for the evaluation.…”

An analysis of evaluation campaigns in ad-hoc medical information retrieval: CLEF eHealth 2013 and 2014

“…proposed methods would in fact significantly improve efficiency over a baseline approach, this increase in effectiveness might not be noted due to missing relevance information in the benchmark data. In particular, it was argued before that experiments with blind relevance feedback or query expansion did not show a significant improvement in performance due to incomplete or missing relevance information [45]. Our analysis shows that for all three strategies investigated, this is clearly not the case (cf.…”

Challenges and Opportunities for Exploring Patient-Level Data

“…In the same vein, data translation also arises as a complex challenge for researchers. In addition to the obvious sense (translating data between multiple languages [ 46 ]), there is the data translation from a low-level free text data to structured information [ 47 , 48 ]. Clinicians' reports traditionally include their notes in free text.…”