LinguaNet: Embedded MT in a Cross-Border Messaging System for European Law Enforcement

Globalisation of crime poses a serious threat to the international community and is a matter of growing concern to law enforcement agencies all over the world. In the combat against international and organized crime, the European Union (EU) has supported a number of research and development projects within the domain of law enforcement focusing on cross-border communication, information extraction and data analysis in a multilingual context as well as terminology and knowledge management. LinguaNet is a case in point and the only project involving a multilingual messaging system. A high level of user involvement was a prominent feature of the project and the resulting software – the LinguaNet system – has gained widespread recognition and usage. The paper gives an overview of the LinguaNet approach as a whole emphasising the temporary experimental embedding of fully automatic MT in a multilingual messaging system. The system is intended for use by professionals with no background in linguistics but in great need of fast and robust communication. One of the conclusions drawn from this experiment is that authoring errors proved to be much more counter-productive than insufficiencies of MT. Another conclusion is that it is preferable to leave it to the recipient of a message to request a machine translation rather than providing it automatically up front. In more general terms, the police liked the approach and reported the need for more message type templates and MT facilities. Finally, the project lead to the formation of a European LinguaNet user group network. This revised version was published online in November 2006 with corrections to the Cover Date.     

Concept-lexeme-syntax triangles: A gateway to interlingual translation

NABU is a large, multilingual Natural Language Processing (NLP) system being developed at MCC for Human Interface applications. Although the NABU project is not considering Machine Translation (MT) as an implementation domain, it is not unreasonable to suppose that, given our multilingual orientation, some MT problems could be ameliorated if not solved by our theoretical approach. This paper addresses the problem of MT via thecognitive interlingua method, focusing on the representation of the lexicon in such a system, and its accommodation of various sources of knowledge for use by both man and machine: notably, in the latter case, morphology, syntax, and semantics. We propose a new theoretical framework — the NABU Word Lattice — as a means of integrating multiple sources of knowledge in a parsimonious fashion conducive to formal interpretation within, and the construction of, an MT system.     

Acquisition of large lexicons for practical knowledge-based MT

Although knowledge-based MT systems have the potential to achieve high translation accuracy, each successful application system requires a large amount of hand-coded lexical knowledge. Systems like KBMT-89 and its descendents have demonstrated how knowledge-based translation can produce good results in technical domains with tractable domain semantics. Nevertheless, the magnitude of the development task for large-scale applications with tens of thousands of domain concepts precludes a purely hand-crafted approach. The current challenge for the next generation of knowledge-based MT systems is to utilize on-line textual resources and corpus analysis software in order to automate the most laborious aspects of the knowledge acquisition process. This partial automation can in turn maximize the productivity of human knowledge engineers and help to make large-scale applications of knowledge-based MT an viable approach. In this paper we discuss the corpus-based knowledge acquisition methodology used in KANT, a knowledge-based translation system for multilingual document production. This methodology can be generalized beyond the KANT interlingua approach for use with any system that requires similar kinds of knowledge.     

Generating Arabic text in multilingual speech-to-speech machine translation framework

The interlingual approach to machine translation (MT) is used successfully in multilingual translation. It aims to achieve the translation task in two independent steps. First, meanings of the source-language sentences are represented in an intermediate language-independent (Interlingua) representation. Then, sentences of the target language are generated from those meaning representations. Arabic natural language processing in general is still underdeveloped and Arabic natural language generation (NLG) is even less developed. In particular, Arabic NLG from Interlinguas was only investigated using template-based approaches. Moreover, tools used for other languages are not easily adaptable to Arabic due to the language complexity at both the morphological and syntactic levels. In this paper, we describe a rule-based generation approach for task-oriented Interlingua-based spoken dialogue that transforms a relatively shallow semantic interlingual representation, called interchange format (IF), into Arabic text that corresponds to the intentions underlying the speaker’s utterances. This approach addresses the handling of the problems of Arabic syntactic structure determination, and Arabic morphological and syntactic generation within the Interlingual MT approach. The generation approach is developed primarily within the framework of the NESPOLE! (NEgotiating through SPOken Language in E-commerce) multilingual speech-to-speech MT project. The IF-to-Arabic generator is implemented in SICStus Prolog. We conducted evaluation experiments using the input and output from the English analyzer that was developed by the NESPOLE! team at Carnegie Mellon University. The results of these experiments were promising and confirmed the ability of the rule-based approach in generating Arabic translation from the Interlingua taken from the travel and tourism domain.     

Multilingual news clustering: Feature translation vs. identification of cognate named entities

In this paper we evaluate the influence of different document representations in the results of multilingual news clustering. We aim at proving whether or not the use of only named entities is a good source of knowledge for multilingual news clustering. We compare two approaches: one based on feature translation, and another based on cognate identification. Our main contribution is using only some categories of cognate named entities like document representation features to perform multilingual news clustering, without the need of translation resources. The results show that the use of cognate named entities, as the only type of features to represent news, leads to good multilingual clustering performance, comparable to the one obtained by using the feature translation approach.     

Recent developments in machine translation

This paper describes developments in the area of machine translation (MT). First, the paper gives an overview of developments in Germany in general; then, special problems are discussed. The system taken as an example is METAL (Machine Translation and Analysis of Natural Language), where recent development work has centered around two main topics. (i) Efforts have been made to make the system really multilingual. The German-to-English prototype had to be expanded, some system components had to be readjusted, and additional problems had to be solved. Currently, analysis and synthesis components for German, English, French, Spanish, and Dutch are under development. All these languages use a common system kernel and a standard interface structure. (ii) The system had to be made user-friendly. This was an even more important task as, up to now, MT systems have not been well accepted by users. METAL tries to be more realistic, and also tries to support the main user interfaces in a much better way than has been done before. This is based on the conviction that there are several parameters which determine the real success of an MT system. It is not just translation quality which is decisive, it is also the integration of an MT system into the whole process of preparing and translating documents.Gregor Thurmair is head of the Linguistics Department at Siemens Nixdorf Information Systems and project leader of the machine translation group, METAL. He is involved in projects in information retrieval (morphological analysis), speech understanding (parsing, semantics) and machine translation (METAL system). He has presented papers on morphology, semantics in speech understanding, transfer problems in MT, and grammar checking.     

Machine Translation of Closed Captions

Traditional Machine Translation (MT) systems are designed to translate documents. In this paper we describe an MT system that translates the closed captions that accompany most North American television broadcasts. This domain has two identifying characteristics. First, the captions themselves have properties quite different from the type of textual input that many MT systems have been designed for. This is due to the fact that captions generally represent speech and hence contain many of the phenomena that characterize spoken language. Second, the operational characteristics of the closed-caption domain are also quite distinctive. Unlike most other translation domains, the translated captions are only one of several sources of information that are available to the user. In addition, the user has limited time to comprehend the translation since captions only appear on the screen for a few seconds. In this paper, we look at some of the theoretical and implementational challenges that these characteristics pose for MT. We present a fully automatic large-scale multilingual MT system, ALTo. Our approach is based on Whitelock's Shake and Bake MT paradigm, which relies heavily on lexical resources. The system currently provides wide-coverage translation from English to Spanish. In addition to discussing the design of the system, we also address the evaluation issues that are associated with this domain and report on our current performance.     

Using target-language information to train part-of-speech taggers for machine translation

Although corpus-based approaches to machine translation (MT) are growing in interest, they are not applicable when the translation involves less-resourced language pairs for which there are no parallel corpora available; in those cases, the rule-based approach is the only applicable solution. Most rule-based MT systems make use of part-of-speech (PoS) taggers to solve the PoS ambiguities in the source-language texts to translate; those MT systems require accurate PoS taggers to produce reliable translations in the target language (TL). The standard statistical approach to PoS ambiguity resolution (or tagging) uses hidden Markov models (HMM) trained in a supervised way from hand-tagged corpora, an expensive resource not always available, or in an unsupervised way through the Baum-Welch expectation-maximization algorithm; both methods use information only from the language being tagged. However, when tagging is considered as an intermediate task for the translation procedure, that is, when the PoS tagger is to be embedded as a module within an MT system, information from the TL can be (unsupervisedly) used in the training phase to increase the translation quality of the whole MT system. This paper presents a method to train HMM-based PoS taggers to be used in MT; the new method uses not only information from the source language (SL), as general-purpose methods do, but also information from the TL and from the remaining modules of the MT system in which the PoS tagger is to be embedded. We find that the translation quality of the MT system embedding a PoS tagger trained in an unsupervised manner through this new method is clearly better than that of the same MT system embedding a PoS tagger trained through the Baum-Welch algorithm, and comparable to that obtained by embedding a PoS tagger trained in a supervised way from hand-tagged corpora.     

System Combination for Machine Translation of Spoken and Written Language

This paper describes an approach for computing a consensus translation from the outputs of multiple machine translation (MT) systems. The consensus translation is computed by weighted majority voting on a confusion network, similarly to the well-established ROVER approach of Fiscus for combining speech recognition hypotheses. To create the confusion network, pairwise word alignments of the original MT hypotheses are learned using an enhanced statistical alignment algorithm that explicitly models word reordering. The context of a whole corpus of automatic translations rather than a single sentence is taken into account in order to achieve high alignment quality. The confusion network is rescored with a special language model, and the consensus translation is extracted as the best path. The proposed system combination approach was evaluated in the framework of the TC-STAR speech translation project. Up to six state-of-the-art statistical phrase-based translation systems from different project partners were combined in the experiments. Significant improvements in translation quality from Spanish to English and from English to Spanish in comparison with the best of the individual MT systems were achieved under official evaluation conditions.      

基于本体的专业机器翻译术语词典研究

在专业机器翻译系统的设计和实现中,要解决的一个关键问题是如何有效地组织面向不同专业领域的专业术语,以及如何根据当前所处理的文本选择相应的术语定义。本文首先分析现有专业机器翻译系统在术语词典组织和建设方面存在的主要问题,以及基于本体(Ontology)的领域知识概念体系的特点;其次,探讨面向专业机器翻译的术语词典研究的几个重要方面,包括通用领域本体的设计、专业术语的描述和向本体的映射、双语或多语MT专业词库的组织和应用等;最后,介绍我们初步已完成的工作,主要包括机器翻译专业领域分类系统设计、专业词典向专业分类系统的映射、ICS标准向专业领域分类系统的映射等。映射实验结果表明,专业领域分类系统对于机器翻译专业词典具有良好的覆盖性。     

Generalizing sampling-based multilingual alignment

Sub-sentential alignment is the process by which multi-word translation units are extracted from sentence-aligned multilingual parallel texts. This process is required, for instance, in the course of training statistical machine translation systems. Standard approaches typically rely on the estimation of several probabilistic models of increasing complexity and on the use of various heuristics, that make it possible to align, first isolated words, then, by extension, groups of words. In this paper, we explore an alternative approach which relies on a much simpler principle: the comparison of occurrence profiles in sub-corpora obtained by sampling. After analyzing the strengths and weaknesses of this approach, we show how to improve the detection of multi-word translation units and evaluate these improvements on machine translation tasks.     

Symbolic-to-statistical hybridization: extending generation-heavy machine translation

The last few years have witnessed an increasing interest in hybridizing surface-based statistical approaches and rule-based symbolic approaches to machine translation (MT). Much of that work is focused on extending statistical MT systems with symbolic knowledge and components. In the brand of hybridization discussed here, we go in the opposite direction: adding statistical bilingual components to a symbolic system. Our base system is Generation-heavy machine translation (GHMT), a primarily symbolic asymmetrical approach that addresses the issue of Interlingual MT resource poverty in source-poor/target-rich language pairs by exploiting symbolic and statistical target-language resources. GHMT’s statistical components are limited to target-language models, which arguably makes it a simple form of a hybrid system. We extend the hybrid nature of GHMT by adding statistical bilingual components. We also describe the details of retargeting it to Arabic–English MT. The morphological richness of Arabic brings several challenges to the hybridization task. We conduct an extensive evaluation of multiple system variants. Our evaluation shows that this new variant of GHMT—a primarily symbolic system extended with monolingual and bilingual statistical components—has a higher degree of grammaticality than a phrase-based statistical MT system, where grammaticality is measured in terms of correct verb-argument realization and long-distance dependency translation.     

Quantitative fine-grained human evaluation of machine translation systems: a case study on English to Croatian

This paper presents a quantitative fine-grained manual evaluation approach to comparing the performance of different machine translation (MT) systems. We build upon the well-established multidimensional quality metrics (MQM) error taxonomy and implement a novel method that assesses whether the differences in performance for MQM error types between different MT systems are statistically significant. We conduct a case study for English-to-Croatian, a language direction that involves translating into a morphologically rich language, for which we compare three MT systems belonging to different paradigms: pure phrase-based, factored phrase-based and neural. First, we design an MQM-compliant error taxonomy tailored to the relevant linguistic phenomena of Slavic languages, which made the annotation process feasible and accurate. Errors in MT outputs were then annotated by two annotators following this taxonomy. Subsequently, we carried out a statistical analysis which showed that the best-performing system (neural) reduces the errors produced by the worst system (pure phrase-based) by more than half (54%). Moreover, we conducted an additional analysis of agreement errors in which we distinguished between short (phrase-level) and long distance (sentence-level) errors. We discovered that phrase-based MT approaches are of limited use for long distance agreement phenomena, for which neural MT was found to be especially effective.     

Toward Web mining of cross-language query translations in digital libraries

This paper proposes an effective query-translation approach that enables a cross-language information retrieval (CLIR) service to be more easily supported in digital library systems that only contain monolingual content. A query-translation engine called LiveTrans is used to process the translation requests of cross-lingual queries from connected digital library systems. To automatically extract translations not covered by standard dictionaries, the engine is developed based on a novel integration of dictionary resources and Web mining approaches, including anchor-text and search-result methods. The engine exploits a broad range of multilingual Web resources used as live bilingual corpora to alleviate translation difficulties. It is shown to be particularly effective for extracting multilingual translation equivalents of query terms containing proper names or new terminology. The obtained results show the feasibility of and great potential for creating English-Chinese CLIR services in existing digital libraries and new applications in cross-language Web searching, although difficulties still remain that need to be investigated further.     

English-Thai Translation: Initial Experiments with a Multiphase Translation System

A multiphase machine translation approach, Generate and Repair Machine Translation (GRMT), is proposed. GRMT is designed to generate accurate translations that focus primarily on retaining the linguistic meaning of the source language sentence. GRMT presently incorporates a limited multilingual translation capability. The central idea behind the GRMT approach is to generate a translationcandidate (TC) by quick and dirty machine translation (QDMT), then investigate the accuracy of that TC by translation candidate evaluation (TCE), and, if necessary, revise the translation in the repair and iterate (RI) phase. To demonstrate the GRMT approach, a translation system that translates from English to Thai has been developed. This paper presents the design characteristics and some experimental results of QDMT and also the initial design, some experiments, and proposed ideas behind TCE and RI.     

Reassessing the proper place of man and machine in translation: a pre-translation scenario

Traditionally, human–machine interaction to reach an improved machine translation (MT) output takes place ex-post and consists of correcting this output. In this work, we investigate other modes of intervention in the MT process. We propose a Pre-Edition protocol that involves: (a) the detection of MT translation difficulties; (b) the resolution of those difficulties by a human translator, who provides their translations (pre-translation); and (c) the integration of the obtained information prior to the automatic translation. This approach can meet individual interaction preferences of certain translators and can be particularly useful for production environments, where more control over output quality is needed. Early resolution of translation difficulties can prevent downstream errors, thus improving the final translation quality “for free”. We show that translation difficulty can be reliably predicted for English for various source units. We demonstrate that the pre-translation information can be successfully exploited by an MT system and that the indirect effects are genuine, accounting for around 16% of the total improvement. We also provide a study of the human effort involved in the resolution process.     

面向篇章机器翻译的英汉翻译单位和翻译模型研究

篇章机器翻译的首要问题是确定翻译单位。基于汉语和英语的语言知识和英汉翻译的实践,该文提出面向篇章机器翻译的基本单位和复合单位的双层单位体系,讨论了这两种单位支持篇章翻译应满足的性质,并据此勾画了篇章机器翻译的拆分、翻译、装配三步模型(PTA模型)。该文提出,汉语篇章机器翻译的复合单位为广义话题结构对应的文本块,基本单位则是根据广义话题结构流水模型得到的话题自足句;英语篇章机器翻译的复合单位为句号句,基本单位为naming-telling小句(NT小句),即指称性成分加上对它的陈述或后修饰成分所构成的小句。该文展示了在这样的翻译单位体系下采用PTA模型的英汉翻译过程实例,规划了面向篇章翻译的英汉小句对齐语料库的建设任务,讨论了PTA模型的可行性。
     

基于核心概念集的多语言Ontology*

随着Web资源的日益丰富,人们需要跨语言的知识共享和信息检索。一个多语言Ontology可以用来刻画不同语言相关领域的知识,克服不同文化和不同语言带来的障碍。对现有的构建多语言Ontology方法进行分析和比较,提出一种基于核心概念集的多语言Ontology的构建方法,用一个独立于特定语言的Ontology以及来自不同自然语言的定义和词汇的同义词集来描述相关领域的概念。用该方法构建的Ontology具有良好的扩展能力、表达能力和推理能力,特别适合分布式环境下大型Ontology的创建。     

A conjoint analysis framework for evaluating user preferences in machine translation

Despite much research on machine translation (MT) evaluation, there is surprisingly little work that directly measures users’ intuitive or emotional preferences regarding different types of MT errors. However, the elicitation and modeling of user preferences is an important prerequisite for research on user adaptation and customization of MT engines. In this paper we explore the use of conjoint analysis as a formal quantitative framework to assess users’ relative preferences for different types of translation errors. We apply our approach to the analysis of MT output from translating public health documents from English into Spanish. Our results indicate that word order errors are clearly the most dispreferred error type, followed by word sense, morphological, and function word errors. The conjoint analysis-based model is able to predict user preferences more accurately than a baseline model that chooses the translation with the fewest errors overall. Additionally we analyze the effect of using a crowd-sourced respondent population versus a sample of domain experts and observe that main preference effects are remarkably stable across the two samples.     

多语言在线机器翻译研究

描述了通过使用外部知识库和基于短语的翻译模型,利用多线程、任务分发的技术实现了一个在线的、高性能的多语言翻译引擎,已初步实现了维汉、哈汉、柯汉三种语言间的翻译。翻译引擎很容易扩展到其他语言对,具有翻译词、短语、句子、文件和网页的功能。