Showing papers on "Computer-assisted translation published in 1998"

Multi-lingual data processing system and system and method for translating text used in computer software utilizing an embedded translator

Abstract: A multi-lingual data processing system can operate in a source language and one or more target languages by automatically translating text, such as application text and system message text. The multi-lingual data processing system includes computer software that can be developed and deployed in a source language and that obtains translated text from one or more translation tables corresponding to the target languages as the computer software is runs on the system. The translation system includes a translation table builder that creates and/or modifies the translation tables by importing translated text and/or by allowing a user to insert translated text. A translation configuration selector allows the user to select translation configuration setting, such as the selected locality. The translation tables include application text translation tables containing translated application text that is obtained by the application programs as each object is created by the application programs. The translation tables also include system message translation tables for each of the target languages containing translated message text such that messages logged by the system can be displayed with the translated message text from the system message translation table corresponding to the selected locality.

A Comparative Study of Query and Document Translation for Cross-Language Information Retrieval

Abstract: Cross-lanuage retrieval systems use queries in one narural language to guide retrieval of documents that might be written in another Acquisition and representation of translation knowledge plays a central role in this process This paper explores the utility of two sources of translation knowledge for cross-language retrieval We have implemented six query translation techniques that use bilingual term lists and one based on direct use of the translation output from an existing machine translation system; these are compared with a document translation technique that uses output from the same machine translation system Average precision measures on a TREC collection suggest that arbitrarily selecting a single dictionary translation is typically no less effective than using every translation in the dictionary, that query translation using a machine translation system can achieve somewhat better effectiveness than simpler techniques, and that document translation may result in further improvements in retrieval effectiveness under some conditions

Target-Text Mediated Interactive Machine Translation

Abstract: The use of Machine Translation as a tool for professional or other highly skilled translators is for the most part currently limited to postediting arrangements in which the translator invokes MT when desired and then manually cleans up the results. A theoretically promising but hitherto largely unsuccessful alternative to postediting for this application is interactive machine translation (IMT), in which the translator and MT system work in tandem. We argue that past failures to make IMT viable as a tool for skilled translators have been the result of an infelicitous mode of interaction rather than any inherent flaw in the idea. As a solution, we propose a new style of IMT in which the target text under construction serves as the medium of communication between an MT system and its user. We describe the design, implementation, and performance of an automatic word completion system for translators which is intended to demonstrate the feasibility of the proposed approach, albeit in a very rudimentary form.

Learning translation templates from examples

Abstract: This paper proposes a mechanism for learning lexical level correspondences between two languages from a set of translated sentence pairs. The proposed mechanism is based on an analogical reasoning between two translation examples. Given two translation examples, the similar parts of the sentences in the source language must correspond to the similar parts of the sentences in the target language. Similarly, the different parts should correspond to the respective parts in the translated sentences. The correspondences between the similarities, and also differences are learned in the form of translation templates. The approach has been implemented and tested on a small training dataset and produced promising results for further investigation.

National language proxy file save and incremental cache translation option for world wide web documents

Abstract: A technique for use in a client-server environment, such as the World Wide Web, for providing a client with the best available translated version of a document requested from a server in a language other than the original language of the document. In response to a request from a client for a document in a secondary language, the technique determines if a version of the document exists in the specified secondary language. The invention supports more than one type of translated version; namely, a machine translation version and a human translation version. Typically, human translation versions are more accurate than machine translation versions. If both translation versions are available and the most recent version is newer than the original document, or if one of the translation versions is available and that version is newer than the original document, the most recent version is sent to the client. If no translation version is available or the available versions are older than the original document, a machine translation version of the original document in the requested language is created and sent to the client.

The Challenges of Cross-Modal Translation: English-to-Sign-Language Translation in the Zardoz System

Abstract: The sign languages used by deaf communities around the world represent a linguistic challenge that natural-language researchers in AI have only recently begun to take up. This challenge is particularly relevant to research in Machine Translation (MT), as natural sign languages have evolved in deaf communities into efficient modes of gestural communication, which differ from English not only in modality but in grammatical structure, exploiting a higher dimensionality of spatial expression. In this paper we describe Zardoz, an on-going AI research system that tackles the cross-modal MT problem, translating English text into fluid sign language. The paper presents an architectural overview of Zardoz, describing its central blackboard organization, the nature of its interlingual representation, and the major components which interact through this blackboard both to analyze the verbal input and generate the corresponding gestural output in one of a number of sign variants.

An interlingua based on domain actions for machine translation of task-oriented dialogues

Abstract: This paper describes an interlingua for spoken language translation that is based on domain actions in the travel planning domain. Domain actions are composed of speech acts (e.g., requestinformation), attributes (e.g., size, price), and objects (e.g., hotel, flight) and can take arguments. Development of the interlingua is guided by a database containing travel dialogues in English, Korean, Japanese, and Italian. There are currently 423 domain actions that cover hotel reservation and transportation. The interlingua will soon be extended to cover tours, tourist attractions, and events. The interlingua is used by the C-STAR speech translation consortium for translating travel planning dialogues in six languages: English, Japanese, German, Korean, Italian, and French. The paper also addresses the role of the interlingua in Carnegie Mellon’s JANUS translation system.

Modeling with Structures in Statistical Machine translation

Abstract: Most statistical machine translation systems employ a word-based alignment model. In this paper we demonstrate that word-based alignment is a major cause of translation errors. We propose a new alignment model based on shallow phrase structures, and the structures can be automatically acquired from parallel corpus. This new model achieved over 10% error reduction for our spoken language translation task.

Dictionary management method and apparatus

Abstract: An information processing apparatus connected to a network to translate a document using a translation knowledge cost-effectively is provided with a dictionary; an inquiry device for transmitting, during a time translating a sentence written in a first language, an inquiry for inquiring the translation knowledge for translating a syntactic unit of the sentence written in the first language containing an unknown word and a correct grammar to the network, after translating other words of the sentence into a second language, when the unknown word that cannot be translated into the second language is recognized during translation of the document; and a translation device for estimating a part of speech for the unknown word to apply a hypothesis-setting rule during translation of the sentence, and for continuing translation of a syntax following the syntactic unit of the sentence, wherein the syntactic unit is translated by using an answer to replace the translation of the sentence when the answer to the inquiry is received.

Manual Queries and Machine Translation in Cross-Language Retrieval and Interactive Retrieval with Cheshire II at TREC-7.

Abstract: For TREC-7, the Berkeley ad-hoc experiments explored more phrase discovery in topics and documents. We utilized Boolean retrieval combined with probabilistic ranking for 17 topics in ad-hoc manual entry. Our cross-language experiments tested 3 di erent widely available machine translation software packages. For language pairs (e.g. German to French) for which no direct machine translation was available we made use of English as a universal intermediate language. For CLIR we also manually reformulated the English topics before doing machine translation, and this elicited a signi cant performance increase for both quad language retrieval and for English against English and French documents. In our Interactive Track entry eight searchers conducted eight searches each, half on the Cheshire II system and the other half on the Zprise system, for a total of 64 searches. Questionnaires were administered to gather information about basic demographic and searching experience, about each search, about each of the systems, and nally, about the user's perceptions of the systems.

Multilingual domain modeling in Twenty-One: automatic creation of a bi-directional translation lexicon from a parallel corpus

Abstract: Within the project Twenty-One, which aims at the effective dissemination of information on ecology and sustainable development, a sytem is developed that supports cross-language information retrieval in any of the four languages Dutch, English, French and German. Knowledge of this application domain is needed to enhance existing translation resources for the purpose of lexical disambiguation. This paper describes an algorithm for the automated acquisition of a translation lexicon from a parallel corpus. New about the presented algorithm is the statistical language model used. Because the algorithm is based on a symmetric translation model it becomes possible to identify one-to-many and many-to-one relations between words of a language pair. We claim that the presented method has two advantages over algorithms that have been published before. Firstly, because the translation model is more powerful, the resulting bilingual lexicon will be more accurate. Secondly, the resulting bilingual lexicon can be used to translate in both directions between a language pair. Different versions of the algorithm were evaluated on the Dutch and English version of the Agenda 21 corpus, which is a UN document on the application domain of sustainable development.

Bricks and Skeletons: Some Ideas for the Near Future of MAHT

Abstract: This paper sets forth some ideas for the evolution of translation tools in the near future. The proposed improvements consist in a closer integration of terminology and sentence databases. In particular, we suggest that bilingual sentence databases (translation memories) could be refined by splitting sentences into large ’’bricks‘‘ of text. We also propose a mechanism through which bilingual sentence databases could be generalized by replacing the known terms by variable placeholders, thus yielding technical sentence ’’skeletons‘‘. This latter idea is the most original and, in our view, the most promising for future implementations. Although these ideas are not yet supported by experiments, we believe that they can be implemented using simple techniques, following the general philosophy that such tools should go as far as possible while remaining robust and useful for the human translator.

Integrating Query Translation and Document Translation in a Cross-language Information Retrieval System

Abstract: Due to the explosive growth of the WWW, very large multilingual textual resources have motivated the researches in Cross-Language Information Retrieval and online Web Machine Translation. In this paper, the integration of language translation and text processing system is proposed to build a multilingual information system. A distributed English-Chinese system on WWW is introduced to illustrate how to integrate query translation, search engines, and web translation system. Since July 1997, more than 46,000 users have accessed our system and about 250,000 English web pages have been translated to pages in Chinese or bilingual English-Chinese versions. And the average satisfaction degree of users at document level is 67.47%.

A Modular Approach to Spoken Language Translation for Large Domains

Abstract: The MT engine of the Janus speech-to-speech translation system is designed around four main principles: 1) an interlingua approach that allows the efficient addition of new languages, 2) the use of semantic grammars that yield low cost high quality translations for limited domains, 3) modular grammars that support easy expansion into new domains, and 4) efficient integration of multiple grammars using multi-domain parse lattices and domain re-scoring. Within the framework of the C-STAR-II speech-to-speech translation effort, these principles are tested against the challenge of providing translation for a number of domains and language pairs with the additional restriction of a common interchange format.

Sakhr Arabic-English Computer-Aided Translation System

Abstract: Automation of the whole translation process using computers and machine translation systems did not fulfill, so far, the ever-growing needs for translation. On the other hand, Computer-aided translation systems have tackled the translation process by first concentrating on those mechanical tasks performed during the translation, then by gradually automating those intellingent (creative) tasks. This has resulted in useful systems that both increase translator's productivity and guarantee better cinsistency across translation jobs. This paper describes Sakhr Cat system which has been specifically designed to support document, web page translation and software localisation for the Arabic-English language pair.

Evaluating resources for query translation in cross-language information retrieval.

Abstract: Our goal is to evaluate the utility of a lexical resource containing Lexical Conceptual Structures LCS for use in cross language information retrieval Our evaluation makes use of a combination of techniques from interlingual machine translation Dorr with conventional information retrieval techniques Oard OardandDorr Given a query in one language we transform the query into the corresponding terms in a second language We evaluate this approach by comparing the resulting retrieval effectiveness with two translation based techniques as well as two techniques for determining the lower and upper bound The main innovation of this work is that it provides a principled framework for controlling translation ambiguity in cross language information retrieval applications Our focus is on the construction of disambiguated target language queries by using verb based entries in our lexicon to construct Lexical Conceptual Structures LCS We view our approach as a crucial step toward the evaluation of the utility of our LCS basedmultilingual lexicon Inaddition this work provides abasis formeasuring the extentto which disambiguation can enhance cross language information retrieval effectiveness

Glossary-Based MT Engines in a Multilingual Analyst‘s Workstation Architecture

Abstract: Fast access to information in different languages is still a major problem for many organizations. We have built a multilingual analyst‘s workstation integrated in the Tipster document management toolkit. The analyst workstation offers to an English-speaking analyst a variety of tools to browse sets of documents in Arabic, Japanese, Spanish and Russian, including a Unicode-based multilingual editor, and a simple machine translation functionality. The Temple project has developed an open multilingual architecture and software support for rapid development of extensible machine translation functionalities. The targeted languages are those for which natural language processing and human resources are scarce or difficult to obtain. The goal is to support rapid development of machine translation functionalities in a very short time with limited resources. Glossary-based machine-translation (GBMT) is used to provide an English gloss of a foreign document. A GBMT system uses a bilingual phrasal dictionary (glossary) to produce a phrase-by-phrase translation. Translation (based on phrase pattern-matching) is fast and accurate regarding the content of the document and browsed documents can be translated almost in real-time. A GBMT system for a language pair is also extremely simple, cheap and fast to develop. Moreover, all language resources used by the system are entirely under the control of the user.

Device and method for machine translation

Abstract: PROBLEM TO BE SOLVED: To obtain a device and a method machine translation which can speedily obtain a translation result necessary and sufficient for a user by performing efficient translation. SOLUTION: A translation management part 102 retrieve the document which is most similar to a current translation object document from documents which are already translated in the past and extracts differences from the document, sentence by sentence. Further, the translation management part 102 instructs a translation engine 103 by using translation control information to translates the original except said repeated parts according to document data of the original. The translation engine 103 performs the machine translation of the document data of the original according to the instruction contents and stores the result in a translation data base 104. An output part 105 outputs the translation result. COPYRIGHT: (C)1999,JPO

Hybrid Approaches to Improvement of Translation Quality in Web-based English-Korean Machine Translation

Abstract: The previous English-Korean MT system that was the transfer-based MT system and applied to only written text enumerated a following brief list of the problems that had not seemed to be easy to solve in the near future: 1) processing of non-continuous idiomatic expressions 2) reduction of too many ambiguities in English syntactic analysis 3) robust processing for failed or illformed sentences 4) selecting correct word correspondence between several alternatives 5) generation of Korean sentence style. The problems can be considered as factors that have influence on the translation quality of machine translation system. This paper describes the symbolic and statistical hybrid approaches to solutions of problems of the previous English-to-Korean machine translation system in terms of the improvement of translation quality. The solutions are now successfully applied to the web-based English-Korean machine translation system "FromTo/EK" which has been developed from 1997.

Sentence-based machine translation for English-Thai

Abstract: This paper proposes a combined method of machine translation in which rule-based techniques and and example-based techniques are applied to produce a new system for English-to-Thai sentence translation. The proposed method has the following four features: (1) analyzing an English sentence into an internal structure based on phrase structure grammar and mapping it onto that Thai equivalent sentence pattern; (2) analyzing the different sections found in input sentence and applying the specific rules for those distinguishing linguistic phenomena; (3) generating output sentences by looking up the bilingual dictionary for the equivalent words; and (4) ranking the possible combinations and eliminating the unacceptable translations. The translated sentences will be stored in a database for retrieving or using as a guide for translation.

Modular language translation system

Abstract: A modular language translation system allows a user of any one of a variety of different user interfaces to be able to send translation requests to and receive responses from any one of a variety of different translation engines. A user of the system familiar with the user interface of a first type of translation system, such as one that translates from Japanese to English, can use that user interface to get translations from the translation engine of a second type of translation system, such as a Russian-to-English system, without having to learn the particularities of the second system and its interface. The user interfaces and the translation engines communicate via a distributed object protocol.

Object oriented translation framework method, apparatus, and program

Abstract: The invention is a translation framework which performs data conversions in an object oriented environment from clipboards, drag, and drop tend entire files and entire objects into compound documents. There are three primary classes: the translator, the translator surrogates, and the translation query. Translators are heavy objects that may include the translator's shared library and supporting libraries. The framework uses light weight surrogate translators as stand-ins for translators. The surrogates do not pull in the translators unless there is a request to do an actual translation. When a translation is requested, the translator must be streamed into the address space of the repeating client. Clients desiring a document object be translated for inclusion in a compound document interact with the framework by using queries to access a translation surrogate.

Time-Constrained Machine Translation

Abstract: This paper defines the class of time-constrained applications: applications in which the user has limited time to process the system output. This class is differentiated from real-time systems, where it is production time rather than comprehension time that is constrained. Examples of time-constrained MT applications include the translation of multi-party dialogue and the translation of closed-captions. The constraints on comprehension time in such systems have significant implications for the system's objectives, its design, and its evaluation. In this paper we outline these challenges and discuss how they have been met in an English-Spanish MT system designed to translate the closed-captions used on television.

Machine translation and the information soup : Third Conference of the Association for Machine Translation in the Americas, AMTA'98, Langhorne, PA, USA, October 28-31, 1998 : proceedings

Abstract: Technical Papers.- A Statistical View on Bilingual Lexicon Extraction: From Parallel Corpora to Non-Parallel Corpora.- Empirical Methods for MT Lexicon Development.- A Modular Approach to Spoken Language Translation for Large Domains.- Enhancing Automatic Acquisition of Thematic Structure in a Large-Scale Lexicon for Mandarin Chinese.- Ordering Translation Templates by Assigning Conifdence Factors.- Quality and Robustness in MT - A Balancing Act.- Parallel Strands: A Preliminary Investigation into Mining the Web for Bilingual Text.- An English-to-Turkish Interlingual MT System.- Rapid Prototyping of Domain-Specific Machine Translation Systems.- Time-Constrained Machine Translation.- An Evaluation of the Multi-engine MT Architecture.- An Ontology-Based Approach to Parsing Turkish Sentences.- Monolingual Translator Workstation.- Fast Document Translation for Cross-Language Information Retrieval.- Machine Translation in Context.- EasyEnglish: Addressing Structural Ambiguity.- Multiple-Subject Constructions in the Multilingual MT-System CAT2.- A Multilingual Procedure for Dictionary-Based Sentence Alignment.- Taxonomy and Lexical Semantics - from the Perspective of Machine Readable Dictionary.- Can Simultaneous Interpretation Help Machine Translation?.- Sentence Analysis Using a Concept Lattice.- Evaluating Language Technologies: The MULTIDOC Approach to Taming the Knowledge Soup.- Integrating Query Translation and Document Translation in a Cross-Language Information Retrieval System.- When Stalhandske Becomes Steelglove.- SYSTRAN on AltaVista A User Study on Real-Time Machine Translation on the Internet.- Making Semantic Interpretation Parser-Independent.- Implementing MT in the Greek Public Sector: A Users' Survey.- Statistical Approach for Korean Analysis: a Method Based on Structural Patterns.- Twisted Pair Grammar: Support for Rapid Development of Machine Translation for Low Density Languages.- A Thematic Hierarchy for Efficient Generation from Lexical-Conceptual Structure.- The LMT Transformational System.- Finding the Right Words: An Analysis of Not-Translated Words in Machine Translation.- Predicting What MT Is Good for: User Judgments and Task Performance.- Reusing Translated Terms to Expand a Multilingual Thesaurus.- Spicing Up the Information Soup: Machine Translation and the Internet.- Revision of Morphological Analysis Errors Through the Person Name Construction Model.- Lexical Choice and Syntactic Generation in a Transfer System Transformations in the New LMT English-German System.- Translation with Finite-State Devices.- Lexical Selection for Cross-Language Applications: Combining LCS with WordNet.- Improving Translation Quality by Manipulating Sentence Length.- Machine Translation among Languages with Transitivity Divergences Using the Causal Relation in the Interlingual Lexicon.- A Comparative Study of Query and Document Translation for Cross-Language Information Retrieval.- Lexicons as Gold: Mining, Embellishment, and Reuse.- System Descriptions.- System Description/Demo of Alis Translation Solutions Overview.- System Demonstration SYSTRAN(R) Enterprise.- Integrating Tools with the Translation Process.- EMIS A Multilingual Information System.- An Open Transfer Translation.- TransEasy: A Chinese-English Machine Translation System Based on Hybrid Approach.- Sakhr Arabic-English Computer-Aided Translation System.- System Description/Demo of Alis Translation Solutions Application: Multilingual Search and Query Expansion.- Logos8 System Description.

Device and method for machine translation

Abstract: PROBLEM TO BE SOLVED: To provide a machine translation device which easily gives translation instruction when an HTML document downloaded from the Internet is translated. SOLUTION: This device is equipped with an HTML downloading means 24 which downloads an HTML document from the Internet, a translating processing means 22 which translates the downloaded HTML document, and a control means 25 which adds a translation flag determining whether or not translation is carried out to link information of the HTML document and displays a translated HTML document. When the translation flag is present, the HTML document which is downloaded from the Internet can automatically be translated, so that a translation is instruction is simplified. COPYRIGHT: (C)1999,JPO

Language Tools at the EC Translation Service: The Theory and the Practice

Abstract: Language technology brings new tools to the translator's desktop: full-text retrieval systems, terminology systems, translation memories and machine translation. European Commission's Translation Service, SdTVista is used for full-text search of reference documents, the powerful aligner in Euramis creates translation memories which are stored in a central Linguistic Resources Database, sentences not found in the memory are automatically sent to the EC Systran machine translation system, and relevant terminology is retrieved in batch mode from the Eurodicautom termbase. The resulting resources are brought together on the Translator's Workbench 2 . These tools are being used extensively today by about one third of SdT translators. While it would be premature to cite global statistics, there are strong indications that substantial gains in productivity can be achieved. The 1996 machine translation survey, the experience of the Translation Workshop between 1996 and 1998, and the 1998 translation memory tools survey confirm this. The crucial factor is the assimilation of translation technology in the workflow of a large organisation — and this constitutes a real challenge for management.

Machine Translation as a Model of Human Translation

Abstract: Machine translation (MT) has generally been seen as a purely engineering enterprise, with virtually no attention paid to whether it provides any kind of psycholinguistic model of human translation (HT). This article investigates whether MT systems offer any kind of parallel to HT, by distinguishing three ways in which MT may simulate HT, namely in terms of input-output relations, knowledge, and processing. It is suggested that MT does not, in general, produce human-like output. And while the knowledge embodied in an MT system is broadly comparable to that of an expert human translator, the former is more compartmentalized, and specifically bilingual knowledge is sparser. Processing is hard to discuss without knowing more about how people translate. However, MT can play a useful role in prompting hypotheses about HT such as whether there is a human analogue of MT's complex transfer.