This paper describes our submission to the WMT20 sentence filtering task. We combine scores from (1) a custom LASER built for each source language, (2) a classifier built to distinguish positive and negative pairs by semantic alignment, and (3) the original scores included in the task devkit. For the mBART finetuning setup, provided by the organizers, our method shows 7% and 5% relative improvement over baseline, in sacreBLEU score on the test set for Pashto and Khmer respectively.
We present a free Japanese-French parallel corpus. It includes 15M aligned segments and is obtained by compiling and filtering several existing resources. In this paper, we describe the existing resources, their quantity and quality, the filtering we applied to improve the quality of the corpus, and the content of the ready-to-use corpus. We also evaluate the usefulness of this corpus and the quality of our filtering by training and evaluating some standard MT systems with it.
In this paper, we describe our submission to the WMT19 low-resource parallel corpus filtering shared task. Our main approach is based on the LASER toolkit (Language-Agnostic SEntence Representations), which uses an encoder-decoder architecture trained on a parallel corpus to obtain multilingual sentence representations. We then use the representations directly to score and filter the noisy parallel sentences without additionally training a scoring function. We contrast our approach to other promising methods and show that LASER yields strong results. Finally, we produce an ensemble of differen…
Peking University（北京大学）；ByteDance AI Lab（字节跳动人工智能实验室）；University of California San Diego（加利福尼亚大学圣迭戈分校）
In this paper, we describe our submissions to the WMT20 shared task on parallel corpus filtering and alignment for low-resource conditions. The task requires the participants to align potential parallel sentence pairs out of the given document pairs, and score them so that low-quality pairs can be filtered. Our system, Volctrans, is made of two modules, i.e., a mining module and a scoring module. Based on the word alignment model, the mining module adopts an iterative mining strategy to extract latent parallel sentences. In the scoring module, an XLM-based scorer provides scores, followed by r…
We present the IIT Bombay English-Hindi Parallel Corpus. The corpus is a compilation of parallel corpora previously available in the public domain as well as new parallel corpora we collected. The corpus contains 1.49 million parallel segments, of which 694k segments were not previously available in the public domain. The corpus has been pre-processed for machine translation, and we report baseline phrase-based SMT and NMT translation results on this corpus. This corpus has been used in two editions of shared tasks at the Workshop on Asian Language Translation (2016 and 2017). The corpus is fr…
We present a new release of the Czech-English parallel corpus CzEng 2.0 consisting of over 2 billion words (2 "gigawords") in each language. The corpus contains document-level information and is filtered with several techniques to lower the amount of noise. In addition to the data in the previous version of CzEng, it contains new authentic and also high-quality synthetic parallel data. CzEng is freely available for research and educational purposes.
This work distinguishes between translated and original text in the UN protocol corpus. By modeling the problem as classification problem, we can achieve up to 95% classification accuracy. We begin by deriving a parallel corpus for different language-pairs annotated for translation direction, and then classify the data by using various feature extraction methods. We compare the different methods as well as the ability to distinguish between translated and original texts in the different languages. The annotated corpus is publicly available.
Center for Information and Language Processing (CIS), LMU Munich（慕尼黑大学信息与语言处理中心（CIS））；Research Center Deutscher Sprachatlas, Philipps University Marburg（马尔堡菲利普大学德国语言地图集研究中心）
With more than 7000 languages worldwide, multilingual natural language processing (NLP) is essential both from an academic and commercial perspective. Researching typological properties of languages is fundamental for progress in multilingual NLP. Examples include assessing language similarity for effective transfer learning, injecting inductive biases into machine learning models or creating resources such as dictionaries and inflection tables. We provide ParCourE, an online tool that allows to browse a word-aligned parallel corpus, covering 1334 languages. We give evidence that this is usefu…
Recent machine translation algorithms mainly rely on parallel corpora. However, since the availability of parallel corpora remains limited, only some resource-rich language pairs can benefit from them. We constructed a parallel corpus for English-Japanese, for which the amount of publicly available parallel corpora is still limited. We constructed the parallel corpus by broadly crawling the web and automatically aligning parallel sentences. Our collected corpus, called JParaCrawl, amassed over 8.7 million sentence pairs. We show how it includes a broader range of domains and how a neural machi…
Text and Speech Technology LTD.（文本与语音技术有限公司）；University of Tehran（德黑兰大学）
In this article, we have introduced the first parallel corpus of Persian with more than 10 other European languages. This article describes primary steps toward preparing a Basic Language Resources Kit (BLARK) for Persian. Up to now, we have proposed morphosyntactic specification of Persian based on EAGLE/MULTEXT guidelines and specific resources of MULTEXT-East. The article introduces Persian Language, with emphasis on its orthography and morphosyntactic features, then a new Part-of-Speech categorization and orthography for Persian in digital environments is proposed. Finally, the corpus and …
With a large amount of parallel data, neural machine translation systems are able to deliver human-level performance for sentence-level translation. However, it is costly to label a large amount of parallel data by humans. In contrast, there is a large-scale of parallel corpus created by humans on the Internet. The major difficulty to utilize them is how to filter them out from the noise website environments. Current parallel data mining methods all require labeled parallel data as the training source. In this paper, we present a pipeline to mine the parallel corpus from the Internet in an uns…
In this work we introduce dual conditional cross-entropy filtering for noisy parallel data. For each sentence pair of the noisy parallel corpus we compute cross-entropy scores according to two inverse translation models trained on clean data. We penalize divergent cross-entropies and weigh the penalty by the cross-entropy average of both models. Sorting or thresholding according to these scores results in better subsets of parallel data. We achieve higher BLEU scores with models trained on parallel data filtered only from Paracrawl than with models trained on clean WMT data. We further evaluat…
ETH Z\" u rich, Department of Computer Science, Switzerland
Swiss German is a dialect continuum whose natively acquired dialects significantly differ from the formal variety of the language. These dialects are mostly used for verbal communication and do not have standard orthography. This has led to a lack of annotated datasets, rendering the use of many NLP methods infeasible. In this paper, we introduce the first annotated parallel corpus of spoken Swiss German across 8 major dialects, plus a Standard German reference. Our goal has been to create and to make available a basic dataset for employing data-driven NLP applications in Swiss German. We pres…
New York University Abu Dhabi（纽约大学阿布扎比分校）；Carnegie Mellon University in Qatar（卡内基梅隆大学卡塔尔分校）
Gender bias in natural language processing (NLP) applications, particularly machine translation, has been receiving increasing attention. Much of the research on this issue has focused on mitigating gender bias in English NLP models and systems. Addressing the problem in poorly resourced, and/or morphologically rich languages has lagged behind, largely due to the lack of datasets and resources. In this paper, we introduce a new corpus for gender identification and rewriting in contexts involving one or two target users (I and/or You) -- first and second grammatical persons with independent gra…
T. J. Watson Research Center, IBM（IBM托马斯·J·沃森研究中心）；University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）
Text style transfer rephrases a text from a source style (e.g., informal) to a target style (e.g., formal) while keeping its original meaning. Despite the success existing works have achieved using a parallel corpus for the two styles, transferring text style has proven significantly more challenging when there is no parallel training corpus. In this paper, we address this challenge by using a reinforcement-learning-based generator-evaluator architecture. Our generator employs an attention-based encoder-decoder to transfer a sentence from the source style to the target style. Our evaluator is …
Vidyasirimedhi Institution of Science and Technology（维迪亚西里梅迪科学技术学院）；pyThaiNLP；Chulalongkorn University（朱拉隆功大学）
The primary objective of our work is to build a large-scale English-Thai dataset for machine translation. We construct an English-Thai machine translation dataset with over 1 million segment pairs, curated from various sources, namely news, Wikipedia articles, SMS messages, task-based dialogs, web-crawled data and government documents. Methodology for gathering data, building parallel texts and removing noisy sentence pairs are presented in a reproducible manner. We train machine translation models based on this dataset. Our models' performance are comparable to that of Google Translation API …
Neural machine translation (NMT) has achieved great successes with large datasets, so NMT is more premised on high-resource languages. This continuously underpins the low resource languages such as Luganda due to the lack of high-quality parallel corpora, so even 'Google translate' does not serve Luganda at the time of this writing. In this paper, we build a parallel corpus with 41,070 pairwise sentences for Luganda and English which is based on three different open-sourced corpora. Then, we train NMT models with hyper-parameter search on the dataset. Experiments gave us a BLEU score of 21.28 …
University of Haifa（海法大学）；Universität des Saarlandes（萨尔大学）
We describe a set of bilingual English--French and English--German parallel corpora in which the direction of translation is accurately and reliably annotated. The corpora are diverse, consisting of parliamentary proceedings, literary works, transcriptions of TED talks and political commentary. They will be instrumental for research of translationese and its applications to (human and machine) translation; specifically, they can be used for the task of translationese identification, a research direction that enjoys a growing interest in recent years. To validate the quality and reliability of …
We present a new, unique and freely available parallel corpus containing European Union (EU) documents of mostly legal nature. It is available in all 20 official EUanguages, with additional documents being available in the languages of the EU candidate countries. The corpus consists of almost 8,000 documents per language, with an average size of nearly 9 million words per language. Pair-wise paragraph alignment information produced by two different aligners (Vanilla and HunAlign) is available for all 190+ language pair combinations. Most texts have been manually classified according to the EUR…
We explore using multilingual document embeddings for nearest neighbor mining of parallel data. Three document-level representations are investigated: (i) document embeddings generated by simply averaging multilingual sentence embeddings; (ii) a neural bag-of-words (BoW) document encoding model; (iii) a hierarchical multilingual document encoder (HiDE) that builds on our sentence-level model. The results show document embeddings derived from sentence-level averaging are surprisingly effective for clean datasets, but suggest models trained hierarchically at the document-level are more effective…
We describe our experience with automatic alignment of sentences in parallel English-Chinese texts. Our report concerns three related topics: (1) progress on the HKUST English-Chinese Parallel Bilingual Corpus; (2) experiments addressing the applicability of Gale & Church's length-based statistical method to the task of alignment involving a non-Indo-European language; and (3) an improved statistical method that also incorporates domain-specific lexical cues.
Bangladesh University of Engineering and Technology (BUET)（孟加拉国工程技术大学）
Despite being the seventh most widely spoken language in the world, Bengali has received much less attention in machine translation literature due to being low in resources. Most publicly available parallel corpora for Bengali are not large enough; and have rather poor quality, mostly because of incorrect sentence alignments resulting from erroneous sentence segmentation, and also because of a high volume of noise present in them. In this work, we build a customized sentence segmenter for Bengali and propose two novel methods for parallel corpus creation on low-resource setups: aligner ensembl…
School of Informatics, University of Edinburgh（爱丁堡大学信息学院）
We present the ELITR ECA corpus, a multilingual corpus derived from publications of the European Court of Auditors. We use automatic translation together with Bleualign to identify parallel sentence pairs in all 506 translation directions. The result is a corpus comprising 264k document pairs and 41.9M sentence pairs.
Enhancing Language Learning through Technology: Introducing a New English-Azerbaijani (Arabic Script) Parallel Corpus
Kartal Ol Research Group（Kartal Ol 研究小组）；Savalan Igidlari Publishing Company（Savalan Igidlari 出版公司）；Research and Development Ultimate（终极研发公司）；Institute of Turkic Studies, Marmara University（马尔马拉大学突厥研究院）；University of Kurdistan（库尔德斯坦大学）；SumerNashr Publishing Company（SumerNashr 出版公司）；Lorestan University（洛雷斯坦大学）；University of Mohaghegh Ardabili（莫哈格·阿达比里大学）；University of Michigan（密歇根大学）
Lurio University（卢里奥大学）；University of Cambridge（剑桥大学）
Major advancement in the performance of machine translation models has been made possible in part thanks to the availability of large-scale parallel corpora. But for most languages in the world, the existence of such corpora is rare. Emakhuwa, a language spoken in Mozambique, is like most African languages low-resource in NLP terms. It lacks both computational and linguistic resources and, to the best of our knowledge, few parallel corpora including Emakhuwa already exist. In this paper we describe the creation of the Emakhuwa-Portuguese parallel corpus, which is a collection of texts from the…
Baidu Research（百度研究院）；Oregon State University（俄勒冈州立大学）
Machine translation requires large amounts of parallel text. While such datasets are abundant in domains such as newswire, they are less accessible in the biomedical domain. Chinese and English are two of the most widely spoken languages, yet to our knowledge a parallel corpus in the biomedical domain does not exist for this language pair. In this study, we develop an effective pipeline to acquire and process an English-Chinese parallel corpus, consisting of about 100,000 sentence pairs and 3,000,000 tokens on each side, from the New England Journal of Medicine (NEJM). We show that training on…
Automated documentation of programming source code and automated code generation from natural language are challenging tasks of both practical and scientific interest. Progress in these areas has been limited by the low availability of parallel corpora of code and natural language descriptions, which tend to be small and constrained to specific domains. In this work we introduce a large and diverse parallel corpus of a hundred thousands Python functions with their documentation strings ("docstrings") generated by scraping open source repositories on GitHub. We describe baseline results for the…
Université François Rabelais Tours（弗朗索瓦·拉伯雷大学图尔分校）；Université de Belgrade（贝尔格莱德大学）
In this paper, we tackle the problem of the translation of proper names. We introduce our hypothesis according to which proper names can be translated more often than most people seem to think. Then, we describe the construction of a parallel multilingual corpus used to illustrate our point. We eventually evaluate both the advantages and limits of this corpus in our study.
We describe an Arabic-Hebrew parallel corpus of TED talks built upon WIT3, the Web inventory that repurposes the original content of the TED website in a way which is more convenient for MT researchers. The benchmark consists of about 2,000 talks, whose subtitles in Arabic and Hebrew have been accurately aligned and rearranged in sentences, for a total of about 3.5M tokens per language. Talks have been partitioned in train, development and test sets similarly in all respects to the MT tasks of the IWSLT 2016 evaluation campaign. In addition to describing the benchmark, we list the problems enc…
We present Multilingual Open Text (MOT), a new multilingual corpus containing text in 44 languages, many of which have limited existing text resources for natural language processing. The first release of the corpus contains over 2.8 million news articles and an additional 1 million short snippets (photo captions, video descriptions, etc.) published between 2001--2022 and collected from Voice of America's news websites. We describe our process for collecting, filtering, and processing the data. The source material is in the public domain, our collection is licensed using a creative commons lic…
University of Oregon（俄勒冈大学）；Princeton University（普林斯顿大学）；Adobe Research（Adobe研究院）
Keyphrase extraction is one of the essential tasks for document understanding in NLP. While the majority of the prior works are dedicated to the formal setting, e.g., books, news or web-blogs, informal texts such as video transcripts are less explored. To address this limitation, in this work we present a novel corpus and method for keyphrase extraction from the transcripts of the videos streamed on the Behance platform. More specifically, in this work, a novel data augmentation is proposed to enrich the model with the background knowledge about the keyphrase extraction task from other domains…
Kyoto University（京都大学）；National Institute of Information and Communications Technology（国立信息与通信技术研究所）
Lectures translation is a case of spoken language translation and there is a lack of publicly available parallel corpora for this purpose. To address this, we examine a language independent framework for parallel corpus mining which is a quick and effective way to mine a parallel corpus from publicly available lectures at Coursera. Our approach determines sentence alignments, relying on machine translation and cosine similarity over continuous-space sentence representations. We also show how to use the resulting corpora in a multistage fine-tuning based domain adaptation for high-quality lectu…
We present hinglishNorm -- a human annotated corpus of Hindi-English code-mixed sentences for text normalization task. Each sentence in the corpus is aligned to its corresponding human annotated normalized form. To the best of our knowledge, there is no corpus of Hindi-English code-mixed sentences for text normalization task that is publicly available. Our work is the first attempt in this direction. The corpus contains 13494 parallel segments. Further, we present baseline normalization results on this corpus. We obtain a Word Error Rate (WER) of 15.55, BiLingual Evaluation Understudy (BLEU) s…
Stanford University（斯坦福大学）；Rakuten Institute of Technology（乐天技术研究所）；Google Brain（谷歌大脑）
In this paper we describe the Japanese-English Subtitle Corpus (JESC). JESC is a large Japanese-English parallel corpus covering the underrepresented domain of conversational dialogue. It consists of more than 3.2 million examples, making it the largest freely available dataset of its kind. The corpus was assembled by crawling and aligning subtitles found on the web. The assembly process incorporates a number of novel preprocessing elements to ensure high monolingual fluency and accurate bilingual alignments. We summarize its contents and evaluate its quality using human experts and baseline m…
Samasāmayik: A Parallel Dataset for Hindi-Sanskrit Machine Translation
Indian Institute of Technology Bombay（印度理工学院孟买分校）；Geakminds Technologies Private Limited（Geakminds技术有限公司）；Indian Institute of Technology Roorkee（印度理工学院罗奥基）
This paper presents the NICT's participation in the WMT18 shared parallel corpus filtering task. The organizers provided 1 billion words German-English corpus crawled from the web as part of the Paracrawl project. This corpus is too noisy to build an acceptable neural machine translation (NMT) system. Using the clean data of the WMT18 shared news translation task, we designed several features and trained a classifier to score each sentence pairs in the noisy data. Finally, we sampled 100 million and 10 million words and built corresponding NMT systems. Empirical results show that our NMT syste…
APPSI-139: A Parallel Corpus of English Application Privacy Policy Summarization and Interpretation
Tianjin University（天津大学）；Zhejiang University（浙江大学）；North University of China（北方工业大学）；Institute of Science Tokyo（东京科学研究所）；The Hong Kong Polytechnic University（香港理工大学）；Hangzhou High-Tech Zone (Binjiang) Institute of Blockchain and Data Security（杭州高新技术区（滨江）区块链与数据安全研究院）
Fair Play in the Newsroom: Actor-Based Filtering Gender Discrimination in Text Corpora
Faculty for Computer Science and Mathematics, Hochschule München University of Applied Sciences（慕尼黑应用科学大学计算机科学与数学学院）；Department of Statistics, LMU Munich（慕尼黑大学统计学系）；Munich Center for Machine Learning (MCML), LMU Munich（慕尼黑大学机器学习中心）
