The cross-lingual lexical substitution task

In this paper we provide an account of the cross-lingual lexical substitution task run as part of SemEval-2010. In this task both annotators (native Spanish speakers, proficient in English) and participating systems had to find Spanish translations for target words in the context of an English sentence. Because only translations of a single lexical unit were required, this task does not necessitate a full blown translation system. This we hope encouraged those working specifically on lexical semantics to participate without a requirement for them to use machine translation software, though they were free to use whatever resources they chose. In this paper we pay particular attention to the resources used by the various participating systems and present analyses to demonstrate the relative strengths of the systems as well as the requirements they have in terms of resources. In addition to the analyses of individual systems we also present the results of a combined system based on voting from the individual systems. We demonstrate that the system produces better results at finding the most frequent translation from the annotators compared to the highest ranked translation provided by individual systems. This supports our other analyses that the systems are heterogeneous, with different strengths and weaknesses.     

基于多粒度对抗训练的鲁棒跨语言对话系统

跨语言对话系统是当前国际研究的热点和难点. 在实际的应用系统搭建中, 通常需要翻译引擎作为不同语言之间对话的桥梁. 然而, 翻译引擎往往是基于不同训练样本构建的, 无论是所在领域, 还是擅长处理语言的特性, 均与对话系统的实际应用需求存在较大的差异, 从而导致整个对话系统的鲁棒性差、响应性能低. 因此, 如何增强跨语言对话系统的鲁棒性对于提升其实用性具有重要的意义. 提出了一种基于多粒度对抗训练的鲁棒跨语言对话系统构建方法. 该方法首先面向机器翻译构建多粒度噪声数据, 分别在词汇、短语和句子层面生成相应的对抗样本, 之后利用多粒度噪声数据和干净数据进行对抗训练, 从而更新对话系统的参数, 进而指导对话系统学习噪声无关的隐层向量表示, 最终达到提升跨语言对话系统性能的目的. 在公开对话数据集上对两种语言的实验表明, 所提出的方法能够显著提升跨语言对话系统的性能, 尤其提升跨语言对话系统的鲁棒性.     

面向神经机器翻译系统的多粒度蜕变测试

机器翻译是利用计算机将一种自然语言转换成另一种自然语言的任务,是人工智能领域研究的热点问题之一.近年来,随着深度学习的发展,基于序列到序列结构的神经机器翻译模型在多种语言对的翻译任务上都取得了超过统计机器翻译模型的效果,并被广泛应用于商用翻译系统中.虽然商用翻译系统的实际应用效果直观表明了神经机器翻译模型性能有很大的提...     

藏汉神经网络机器翻译研究

神经网络机器翻译是最近几年提出的机器翻译方法,在多数语言对上逐渐超过了统计机器翻译方法,成为当前机器翻译研究前沿热点。该文在藏汉语对上进行了基于注意力的神经网络机器翻译的实验,并采用迁移学习方法缓解藏汉平行语料数量不足问题。实验结果显示,该文提出的迁移学习方法简单有效,相比短语统计机器翻译方法,提高了三个BLEU值。从译文分析中可以看出藏汉神经网络机器翻译的译文比较流畅,远距离调序能力较强,同时也存在过度翻译、翻译不充分、翻译忠实度较低等神经网络机器翻译的共同不足之处。     

Scheduling Task Parallel Applications for Rapid Turnaround on Enterprise Desktop Grids

Desktop Grids are popular platforms for high throughput applications, but due to their inherent resource volatility it is difficult to exploit them for applications that require rapid turnaround. Efficient desktop Grid execution of short-lived applications is an attractive proposition and we claim that it is achievable via intelligent resource selection. We propose three general techniques for resource selection: resource prioritization, resource exclusion, and task duplication. We use these techniques to instantiate several scheduling heuristics. We evaluate these heuristics through trace-driven simulations of four representative desktop Grid configurations. We find that ranking desktop resources according to their clock rates, without taking into account their availability history, is surprisingly effective in practice. Our main result is that a heuristic that uses the appropriate combination of resource prioritization, resource exclusion, and task replication can achieve performance within a factor of 1.7 of optimal in practice.     

Improving neural sentence alignment with word translation

Sentence alignment is a basic task in natural lan-guage processing which aims to extract high-quality paral-lel sentences automatically.Motivated by the observation that aligned sentence pairs contain a larger number of aligned words than unaligned ones,we treat word translation as one of the most useful external knowledge.In this paper,we show how to explicitly integrate word translation into neural sentence alignment.Specifically,this paper proposes three cross-lingual encoders to incorporate word translation:1)Mixed Encoder that learns words and their translation annotation vectors over sequences where words and their translations are mixed alterma-tively;2)Factored Encoder that views word translations as fea-tures and encodes words and their translations by concatenating their embeddings;and 3)Gated Encoder that uses gate mechanism to selectively control the amount of word translations moving forward.Experimentation on NIST MT and Opensub-titles Chinese-English datasets on both non-monotonicity and monotonicity scenarios demonstrates that all the proposed encoders significantly improve sentence alignment performance.     

机器翻译译文质量估计综述

机器翻译译文质量估计(Quality Estimation, QE)是指在不需要人工参考译文的条件下,估计机器翻译系统产生的译文的质量,对机器翻译研究和应用具有很重要的价值。机器翻译译文质量估计经过最近几年的发展,取得了丰富的研究成果。该文首先介绍了机器翻译译文质量估计的背景与意义;然后详细介绍了句子级QE、单词级QE、文档级QE的具体任务目标、评价指标等内容,进一步概括了QE方法发展的三个阶段：基于特征工程和机器学习的QE方法阶段,基于深度学习的QE方法阶段,融入预训练模型的QE方法阶段,并介绍了每一阶段中的代表性研究工作;最后分析了目前的研究现状及不足,并对未来QE方法的研究及发展方向进行了展望。     

The scaling problem in the pattern recognition approach to machine translation

Statistical machine translation (SMT) has proven to be an interesting pattern recognition framework for automatically building machine translations systems from available parallel corpora. In the last few years, research in SMT has been characterized by two significant advances. First, the popularization of the so called phrase-based statistical translation models, which allows to incorporate local contextual information to the translation models. Second, the availability of larger and larger parallel corpora, which are composed of millions of sentence pairs, and tens of millions of running words. Since phrase-based models basically consists in statistical dictionaries of phrase pairs, their estimation from very large corpora is a very costly task that yields a huge number of parameters which are to be stored in memory. The handling of millions of model parameters and a similar number of training samples have become a bottleneck in the field of SMT, as well as in other well-known pattern recognition tasks such as speech recognition or handwritten recognition, just to name a few. In this paper, we propose a general framework that deals with the scaling problem in SMT without introducing significant time overhead by means of the combination of different scaling techniques. This new framework is based on the use of counts instead of probabilities, and on the concept of cache memory.     

基于中心语块扩展的短语对齐

短语等价对在词典编纂、机器翻译和跨语言信息检索中有着广泛的应用．提出了一种新的短语对齐方法,使用可信度较高的词典对齐结果来抽取源语言短语的译文中心语块,依据译文扩展可信度来确定源语言短语的译文统计边界．从译文中心语块出发,结合译文统计边界生成源语言短语的所有候选译文．对候选译文进行评价,从中选出最可靠的译文．同时利用贪心算法消除源语言短语译文边界之间的交叉冲突．实验结果表明,所提出的方法在开放测试中其正确率达到了82．76％,性能好于其他方法．     

引入源端信息的机器译文自动评价方法研究

机器译文自动评价是机器翻译中的一个重要任务。针对目前译文自动评价中完全忽略源语言句子信息,仅利用人工参考译文度量翻译质量的不足,该文提出了引入源语言句子信息的机器译文自动评价方法: 从机器译文与其源语言句子组成的二元组中提取描述翻译质量的质量向量,并将其与基于语境词向量的译文自动评价方法利用深度神经网络进行融合。在WMT-19译文自动评价任务数据集上的实验结果表明,该文所提出的方法能有效增强机器译文自动评价与人工评价的相关性。深入的实验分析进一步揭示了源语言句子信息在译文自动评价中发挥着重要作用。     

Predicting sentence translation quality using extrinsic and language independent features

We develop a top performing model for automatic, accurate, and language independent prediction of sentence-level statistical machine translation (SMT) quality with or without looking at the translation outputs. We derive various feature functions measuring the closeness of a given test sentence to the training data and the difficulty of translating the sentence. We describe mono feature functions that are based on statistics of only one side of the parallel training corpora and duo feature functions that incorporate statistics involving both source and target sides of the training data. Overall, we describe novel, language independent, and SMT system extrinsic features for predicting the SMT performance, which also rank high during feature ranking evaluations. We experiment with different learning settings, with or without looking at the translations, which help differentiate the contribution of different feature sets. We apply partial least squares and feature subset selection, both of which improve the results and we present ranking of the top features selected for each learning setting, providing an exhaustive analysis of the extrinsic features used. We show that by just looking at the test source sentences and not using the translation outputs at all, we can achieve better performance than a baseline system using SMT model dependent features that generated the translations. Furthermore, our prediction system is able to achieve the $2$ nd best performance overall according to the official results of the quality estimation task (QET) challenge when also looking at the translation outputs. Our representation and features achieve the top performance in QET among the models using the SVR learning model.     

基于预训练的无监督神经机器翻译模型研究

依赖于大规模的平行语料库,神经机器翻译在某些语言对上已经取得了巨大的成功。无监督神经机器翻译UNMT又在一定程度上解决了高质量平行语料库难以获取的问题。最近的研究表明,跨语言模型预训练能够显著提高UNMT的翻译性能,其使用大规模的单语语料库在跨语言场景中对深层次上下文信息进行建模,获得了显著的效果。进一步探究基于跨语言预训练的UNMT,提出了几种改进模型训练的方法,针对在预训练之后UNMT模型参数初始化质量不平衡的问题,提出二次预训练语言模型和利用预训练模型的自注意力机制层优化UNMT模型的上下文注意力机制层2种方法。同时,针对UNMT中反向翻译方法缺乏指导的问题,尝试将Teacher-Student框架融入到UNMT的任务中。实验结果表明,在不同语言对上与基准系统相比,本文的方法最高取得了0.8～2.08个百分点的双语互译评估(BLEU)值的提升。     

基于依存图网络的汉越神经机器翻译方法

汉越神经机器翻译是典型的低资源翻译任务，由于缺少大规模的平行语料，可能导致模型对双语句法差异学习不充分，翻译效果不佳。句法的依存关系对译文生成有一定的指导和约束作用，因此，该文提出一种基于依存图网络的汉越神经机器翻译方法。该方法利用依存句法关系构建依存图网络并融入神经机器翻译模型中，在Transformer模型框架下，引入一个图编码器，对源语言的依存结构图进行向量化编码，利用多头注意力机制，将向量化的依存图结构编码融入到序列编码中，在解码时利用该结构编码和序列编码一起指导模型解码生成译文。实验结果表明，在汉越翻译任务中，融入依存句法图可以提升翻译模型的性能。     

跨语言词向量研究综述

随着人们对互联网多语言信息需求的日益增长,跨语言词向量已成为一项重要的基础工具,并成功应用到机器翻译、信息检索、文本情感分析等自然语言处理领域。跨语言词向量是单语词向量的一种自然扩展,词的跨语言表示通过将不同的语言映射到一个共享的低维向量空间,在不同语言间进行知识转移,从而在多语言环境下对词义进行准确捕捉。近几年跨语言词向量模型的研究成果比较丰富,研究者们提出了较多生成跨语言词向量的方法。该文通过对现有的跨语言词向量模型研究的文献回顾,综合论述了近年来跨语言词向量模型、方法、技术的发展。按照词向量训练方法的不同,将其分为有监督学习、无监督学习和半监督学习三类方法,并对各类训练方法的原理和代表性研究进行总结以及详细的比较;最后概述了跨语言词向量的评估及应用,并分析了所面临的挑战和未来的发展方向。     

Resource allocation in grid computing

Grid computing, in which a network of computers is integrated to create a very fast virtual computer, is becoming ever more prevalent. Examples include the TeraGrid and Planet-lab.org, as well as applications on the existing Internet that take advantage of unused computing and storage capacity of idle desktop machines, such as Kazaa, SETI@home, Climateprediction.net, and Einstein@home. Grid computing permits a network of computers to act as a very fast virtual computer. With many alternative computers available, each with varying extra capacity, and each of which may connect or disconnect from the grid at any time, it may make sense to send the same task to more than one computer. The application can then use the output of whichever computer finishes the task first. Thus, the important issue of the dynamic assignment of tasks to individual computers is complicated in grid computing by the option of assigning multiple copies of the same task to different computers. We show that under fairly mild and often reasonable conditions, maximizing task replication stochastically maximizes the number of task completions by any time. That is, it is better to do the same task on as many computers as possible, rather than assigning different tasks to individual computers. We show maximal task replication is optimal when tasks have identical size and processing times have a NWU (New Worse than Used; defined later) distribution. Computers may be heterogeneous and their speeds may vary randomly, as is the case in grid computing environments. We also show that maximal task replication, along with a c μ rule, stochastically maximizes the successful task completion process when task processing times are exponential and depend on both the task and computer, and tasks have different probabilities of completing successfully.     

A configurable translation-based cross-lingual ontology mapping system to adjust mapping outcomes

Ontologies are widely considered as the building blocks of the semantic web, and with them, comes the data interoperability issue. As ontologies are not necessarily always labelled in the same natural language, one way to achieve semantic interoperability is by means of cross-lingual ontology mapping. Translation techniques are often used as an intermediate step to translate the conceptual labels within an ontology. This approach essentially removes the natural language barrier in the mapping environment and enables the application of monolingual ontology mapping tools. This paper shows that the key to this translation-based approach to cross-lingual ontology mapping lies with selecting appropriate ontology label translations in a given mapping context. Appropriateness of the translations in the context of cross-lingual ontology mapping differs from the ontology localisation point of view, as the former aims to generate correct mappings whereas the latter aims to adapt specifications of conceptualisations to target communities. This paper further demonstrates that the mapping outcome using the translation-based cross-lingual ontology mapping approach is conditioned on the translations selected for the intermediate label translation step. In particular, this paper presents the design, implementation and evaluation of a novel cross-lingual ontology mapping system: SOCOM++. SOCOM++ provides configurable properties that can be manipulated by a user in the process of selecting label translations in an effort to adjust the subsequent mapping outcome. It is shown through the evaluation that for the same pair of ontologies, the mappings between them can be adjusted by tuning the translations for the ontology labels. This finding is not yet shown in the previous research.     

Collocation Dictionary Optimization Using WordNet and k-Nearest Neighbor Learning

In machine translation, collocation dictionaries are important for selecting accurate target words. However, if the dictionary size is too large it can decrease the efficiency of translation. This paper presents a method to develop a compact collocation dictionary for transitive verb–object pairs in English–Korean machine translation without losing translation accuracy. We use WordNet to calculate the semantic distance between words, and k-nearestneighbor learning to select the translations. The entries in the dictionary are minimized to balance the trade-off between translation accuracy and time. We have performed several experiments on a selected set of verbs extracted from a raw corpus of over 3 million words. The results show that in real-time translation environments the size of a collocation dictionary can be reduced up to 40% of its original size without significant decrease in its accuracy.     

A Technical Word- and Term-Translation Aid Using Noisy Parallel Corpora across Language Groups

Technical-term translation represents one of the most difficult tasks for human translators since (1) most translators are not familiar with terms and domain-specific terminology and (2) such terms are not adequately covered by printed dictionaries. This paper describes an algorithm for translating technical words and terms from noisy parallel corpora across language groups. Given any word which is part of a technical term in the source language, the algorithm produces a ranked candidate match for it in the target language. Potential translations for the term are compiled from the matched words and are also ranked. We show how this ranked list helps translators in technical-term translation. Most algorithms for lexical and term translation focus on Indo-European language pairs, and most use a sentence-aligned clean parallel corpus without insertion, deletion or OCR noise. Our algorithm is language- and character-set-independent, and is robust to noise in the corpus. We show how our algorithm requires minimum preprocessing and is able to obtain technical-word translations without sentence-boundary identification or sentence alignment, from the English–Japanese awk manual corpus with noise arising from text insertions or deletions and on the English–Chinese HKUST bilingual corpus. We obtain a precision of 55.35% from the awk corpus for word translation including rare words, counting only the best candidate and direct translations. Translation precision of the best-candidate translation is 89.93% from the HKUST corpus. Potential term translations produced by the program help bilingual speakers to get a 47% improvement in translating technical terms.     

网格经济模型中基于信任机制的调度算法

在网格环境中使用经济学原理进行资源管理和调度是网格从理论研究走向实际应用的有效途径之一。本文在Buyya设计的GRACE网格资源管理框架下,提出一种基于微分方程的信任值量化计算公式:信任=直接信任8声誉,在此基础上建立基于行为的网格信任机制。根据应用环境的不同需求,对网格经济模型调度算法(DBC)进行改进,分别提出了以时间优化、成本优化和时间成本折衷优化为目的的网格信任调度算法(TrustDBC)。理论分析及模拟实验结果表明,本文算法性能明显优于相应的未考虑信任的调度算法。     

Personalising speech-to-speech translation: Unsupervised cross-lingual speaker adaptation for HMM-based speech synthesis

In this paper we present results of unsupervised cross-lingual speaker adaptation applied to text-to-speech synthesis. The application of our research is the personalisation of speech-to-speech translation in which we employ a HMM statistical framework for both speech recognition and synthesis. This framework provides a logical mechanism to adapt synthesised speech output to the voice of the user by way of speech recognition. In this work we present results of several different unsupervised and cross-lingual adaptation approaches as well as an end-to-end speaker adaptive speech-to-speech translation system. Our experiments show that we can successfully apply speaker adaptation in both unsupervised and cross-lingual scenarios and our proposed algorithms seem to generalise well for several language pairs. We also discuss important future directions including the need for better evaluation metrics.