Abstract:One of the most promising approaches to machine translation consists in formulating the problem by means of a pattern recognition approach. By doing so, there are some tasks in which online adaptation is needed in order to adapt the system to changing scenarios. In the present work, we perform an exhaustive comparison of four online learning algorithms when combined with two adaptation strategies for the task of online adaptation in statistical machine translation. Two of these algorithms are already well-know… Show more
“…Incremental MT learning has been investigated several times, usually starting from no data (Barrachina et al, 2009;Ortiz-Martínez et al, 2010), via simulated post-editing (Martínez-Gómez et al, 2012;Denkowski et al, 2014a), or via re-ranking (Wäschle et al, 2013). No previous experiments combined large-scale baselines, full re-tuning of the model weights, and HTER optimization.…”
Analyses of computer aided translation typically focus on either frontend interfaces and human effort, or backend translation and machine learnability of corrections. However, this distinction is artificial in practice since the frontend and backend must work in concert. We present the first holistic, quantitative evaluation of these issues by contrasting two assistive modes: postediting and interactive machine translation (MT). We describe a new translator interface, extensive modifications to a phrasebased MT system, and a novel objective function for re-tuning to human corrections. Evaluation with professional bilingual translators shows that post-edit is faster than interactive at the cost of translation quality for French-English and EnglishGerman. However, re-tuning the MT system to interactive output leads to larger, statistically significant reductions in HTER versus re-tuning to post-edit. Analysis shows that tuning directly to HTER results in fine-grained corrections to subsequent machine output.
“…Incremental MT learning has been investigated several times, usually starting from no data (Barrachina et al, 2009;Ortiz-Martínez et al, 2010), via simulated post-editing (Martínez-Gómez et al, 2012;Denkowski et al, 2014a), or via re-ranking (Wäschle et al, 2013). No previous experiments combined large-scale baselines, full re-tuning of the model weights, and HTER optimization.…”
Analyses of computer aided translation typically focus on either frontend interfaces and human effort, or backend translation and machine learnability of corrections. However, this distinction is artificial in practice since the frontend and backend must work in concert. We present the first holistic, quantitative evaluation of these issues by contrasting two assistive modes: postediting and interactive machine translation (MT). We describe a new translator interface, extensive modifications to a phrasebased MT system, and a novel objective function for re-tuning to human corrections. Evaluation with professional bilingual translators shows that post-edit is faster than interactive at the cost of translation quality for French-English and EnglishGerman. However, re-tuning the MT system to interactive output leads to larger, statistically significant reductions in HTER versus re-tuning to post-edit. Analysis shows that tuning directly to HTER results in fine-grained corrections to subsequent machine output.
“…To overcome this problem, several incremental alignment models have been proposed in the literature (Levenberg, Callison-Burch, & Osborne, 2010). With the exception of the stream-based translation approach, which adds or updates the original TM scores according to the new material (Ortiz-Martínez, García-Varea, & Casacuberta, 2010;Martínez-Gómez, Sanchis-Trilles, & Casacuberta, 2012;Mathur et al, 2013), the adaptation step is usually carried out by creating specific translation tables from the edited translations (using the standard phrase-extraction and phrase-scoring algorithms) and then combining them with the original translation tables. It is important to note that most of the work on incremental adaptation has been tested in the scenarios where references are used instead of UEs.…”
Section: Related Workmentioning
confidence: 99%
“…However, only simulated data, using references instead of actual human UEs, were considered in this work. Ortiz-Martínez et al (2010) and Martínez-Gómez et al (2012) applied an incremental version of the Expectation Maximization (EM) algorithm (Neal & Hinton, 1998) that minimizes an error function with small sequences of mini-batched data. This paradigm is commonly known as stream-based translation, as small portions of data are processed over time.…”
In this article we present a three-step methodology for dynamically improving a statistical machine translation (SMT) system by incorporating human feedback in the form of free edits on the system translations. We target at feedback provided by casual users, which is typically error-prone. Thus, we first propose a filtering step to automatically identify the better user-edited translations and discard the useless ones. A second step produces a pivot-based alignment between source and user-edited sentences, focusing on the errors made by the system. Finally, a third step produces a new translation model and combines it linearly with the one from the original system. We perform a thorough evaluation on a real-world dataset collected from the Reverso.net translation service and show that every step in our methodology contributes significantly to improve a general purpose SMT system. Interestingly, the quality improvement is not only due to the increase of lexical coverage, but to a better lexical selection, reordering, and morphology. Finally, we show the robustness of the methodology by applying it to a different scenario, in which the new examples come from an automatically Web-crawled parallel corpus. Using exactly the same architecture and models provides again a significant improvement of the translation quality of a general purpose baseline SMT system.
“…On the MT system side, research on adaptive approaches tailored to interactive SMT and CAT scenarios explored the online learning protocol (Littlestone, 1988) to improve various aspects of the decoding process (Cesa-Bianchi et al, 2008;Ortiz-Martínez et al, 2010;Martínez-Gómez et al, 2011;Martínez-Gómez et al, 2012;Mathur et al, 2013;Bertoldi et al, 2013).…”
The automatic estimation of machine translation (MT) output quality is a hard task in which the selection of the appropriate algorithm and the most predictive features over reasonably sized training sets plays a crucial role. When moving from controlled lab evaluations to real-life scenarios the task becomes even harder. For current MT quality estimation (QE) systems, additional complexity comes from the difficulty to model user and domain changes. Indeed, the instability of the systems with respect to data coming from different distributions calls for adaptive solutions that react to new operating conditions. To tackle this issue we propose an online framework for adaptive QE that targets reactivity and robustness to user and domain changes. Contrastive experiments in different testing conditions involving user and domain changes demonstrate the effectiveness of our approach.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.