The Prague Bulletin of Mathematical Linguistics 

About: The Prague Bulletin of Mathematical Linguistics is an academic journal published by De Gruyter Open. The journal publishes majorly in the area(s): Machine translation & Treebank. It has an ISSN identifier of 0032-6585. It is also open access. Over the lifetime, 317 publications have been published receiving 4325 citations.

A Productivity Test of Statistical Machine Translation Post-Editing in a Typical Localisation Context

Abstract: We evaluated the productivity increase of statistical MT post-editing as compared to traditional translation in a two-day test involving twelve participants translating from English to French, Italian, German, and Spanish. The test setup followed an empirical methodology. A random subset of the entire new content produced in our company during a given year was translated with statistical MT engines trained on data from the previous year. The translation environment recorded translation and post-editing times for each sentence. The results show a productivity increase for each participant, with significant variance across inviduals.

Training Tips for the Transformer Model

Abstract: This article describes our experiments in neural machine translation using the recent Tensor2Tensor framework and the Transformer sequence-to-sequence model (Vaswani et al., 2017). We examine some of the critical parameters that affect the final translation quality, memory usage, training stability and training time, concluding each experiment with a set of recommendations for fellow researchers. In addition to confirming the general mantra "more data and larger models", we address scaling to multiple GPUs and provide practical tips for improved training regarding batch size, learning rate, warmup steps, maximum sentence length and checkpoint averaging. We hope that our observations will allow others to get better results given their particular hardware and data constraints.

Is Neural Machine Translation the New State of the Art

Abstract: This paper discusses neural machine translation (NMT), a new paradigm in the MT field, comparing the quality of NMT systems with statistical MT by describing three studies using automatic and human evaluation methods. Automatic evaluation results presented for NMT are very promising, however human evaluations show mixed results. We report increases in fluency but inconsistent results for adequacy and post-editing effort. NMT undoubtedly represents a step forward for the MT field, but one that the community should be careful not to oversell.

Z-MERT: A Fully Configurable Open Source Tool for Minimum Error Rate Training of Machine Translation Systems

Abstract: We introduce Z-MERT, a soware tool for minimum error rate training of machine translation systems (Och, 2003). In addition to being an open source tool that is extremely easy to compile and run, Z-MERT is also agnostic regarding the evaluation metric, fully configurable, and requires no modification to work with any decoder. We describe Z-MERT and review its features, and report the results of a series of experiments that examine the tool’s runtime. We establish that Z-MERT is extremely efficient, making it well-suited for time-sensitive pipelines. e experiments also provide an insight into the tool’s runtime in terms of several variables (size of the development set, size of produced N-best lists, etc).

Efficient Word Alignment with Markov Chain Monte Carlo

Abstract: We present efmaral, a new system for efficient and accurate word alignment using a Bayesian model with Markov Chain Monte Carlo (MCMC) inference. Through careful selection of data structures and mo ...