The Future of Translation Technology Towards a World without Babel (3)

Чтобы посмотреть этот PDF файл с форматированием и разметкой, скачайте его и откройте на своем компьютере.

The Future of Translation
Technology has revolutionized the Þ eld of translation, bringing drastic changes
to the way translation is studied and done. To an average user, technology is
beyond a systemÕs interface to see what is at work and what should be done to
make it work more efÞ
ciently. This book is both macroscopic and microscopic
in approach: macroscopic as it adopts a holistic orientation when outlining the
development of translation technology in the last forty years, organizing concepts

Routledge studies in translation technology

1 The Future of Translation Technology
Towards a World without Babel
For a full list of titles in this series, please visit
The Future of Translation
Towards a World without Babel
Chan Sin-wai
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
711 Third Avenue, New York, NY 10017

Routledge is an imprint of the Taylor & Francis Group,
an informa business

The right of Chan Sin-wai to be identiÞ ed as author of this work
has been asserted by him in accordance with sections 77 and 78
All rights reserved. No part of this book may be reprinted or
reproduced or utilized in any form or by any electronic,
mechanical, or other means, now known or hereafter invented,
including photocopying and recording, or in any information
1 The development of translation technology: 1967Ð2014
2 Major concepts in computer-aided translation
3 Functions in computer-aided translation systems
4 Computer-aided translation: Free and paid systems
2.1 Screenshot of Dashboard of SDL-Trados 2015
2.2 Screenshot of SDL-Trados: ÔProgress of
Individual ProjectsÕ
2.1 Statistics of Languages Support by Seven CAT Systems
3.1 Project Wizards of Nine CAT Systems
3.2 Designations of Terminology Databases of
Nine CAT Systems
3.3 Designations of Translation Memory Databases of
Nine CAT Systems
3.4 Editing Environments of Nine CAT Systems
who read through the manuscript and made many useful suggestions. For the
publication of this book and creation of the Routledge Translation Technology
Studies, my gratitude goes to Miss Christina Low, Commissioning Editor of
Figure ÒTranslation Editor of MT2007Ó, featured in Chapter 4 . Used with
kind permission of Andrew Manson.
Figure ÒEditor of OmegaTÓ, featured in Chapter 4 . Used with kind permis-
sion of OmegaT (Marc Prior).
Figures ÒSimilis ManagerÓ and ÒSimilis: Statistics of Translation TasksÓ featured
The history of translation technology, or more speciÞ
cally computer-aided
translation (CAT), is short, but its development has been fast (Chan 2015: 3).
It is generally recognized that the failure of machine translation in the 1960s
as a result of the infamous ALPAC report (1966) led to the emergence of
computer-aided translation. The development of computer-aided translation in
the course of the last forty-seven years, from its beginning in 1967 to 2014,
can be divided into four periods. The Þ rst period, which goes from 1967 to
The development of translation technology
book on machine translation (Locke and Booth 1955). In 1954, Leon Dostert
to John Hutchins, the concept of translation memory can be traced back to
The development of translation technology
of this generation of computer-aided translation systems was that sentences with
full matching were very small in number, minimizing the reusability of the
translation memory and the role of the translation memory database (Wang
Some researchers around 1980 began to collect and store translation samples
with the intention of redeploying and sharing their translation resources. Con-
strained by the limitations of computer hardware (such as limited storage space),
the algorithms for bilingual data alignment, translation memory technology was
forced to remain in a stage of exploration. As a result, a truly commercial
computer-aided translation system did not emerge during the sixteen years of
this period, and, therefore, translation technology did not have an impact on
the translation practice and translation industry (Zachary 1979: 13Ð28).
It can be observed that during this early period of computer-aided translation,
all companies in the Þ eld either were established or operated in Europe. This
Eurocentric phenomenon was bound to change in the next period.
System commercialization
The commercialization of computer-aided translation systems began in 1988,
when Eiichiro Sumita and Yutaka Tsutsumi of the Japanese branch of IBM
The development of translation technology
featured the modules that are standard features of todayÕs computer-aided
technology, which has been at the heart of all PROMT products. Later, they
began to provide a full range of translation solutions: MT systems and services,
The development of translation technology
as the Þ rst of its kind. Version 1.1 followed soon afterwards, incorporating
several performance improvements and an integrated alignment tool (at a time
An-Nakel Al-Arabi, with features that comprised MT, customized dictionaries,
and translation memory. Because of its deep sentence analysis and semantic
connections, An-Nakeel Al-Arabi could learn new rules and knowledge. CIMOS
The development of translation technology
States, Trados 5.5 (Trados Corporate Translation Solutionª) was released. In
Europe, and more speciÞ cally, in the United Kingdom, SDL International
launched its new SDLX Translation Suite 4 and, later this year, released the
elite version of the suite. The SDLX Translation Suite featured a modular
architecture consisting of Þ ve to eight components: SDL Project Wizard, SDL
Align, SDL Maintain, SDL Edit, and SDL TermBase, present in all versions;
and SDL Analyse, SDL Apply, and SDLX AutoTrans, only present in the Pro-
The development of translation technology
As aforementioned, PROMT released a new version, PROMT Expert, inte-
The support of translation of more languages
Translation memory is supposed to be language-independent, but computer-
aided translation systems developed in the early 1990s deÞ nitely did not support
all languages. In 1992, TranslatorÕs Workbench Editor, for instance, could handle
merely Þ ve European languages, namely German, English, French, Italian, and
Spanish; whereas IBM Translation Manager/2 already supported 19 languages,
The development of translation technology
This situation has offered a wider range of choices for buyers to acquire systems
with different packages, functions, operating systems, and prices.
One of the most signiÞ cant changes in this period is the addition of new
computer-aided translation companies in countries other than those mentioned
above. Hungary is a typical example. In 2004, Kilgray Translation Technolo-
gies was established by three Hungarian language technologists. The name of
the company was made up of the foundersÕ surnames: Kis Bal‡zs (KI), Lengyel
Istv‡n (L), and Ugray G‡bor (GRAY). Shortly after, in 2005, the company
rst version of MemoQ, an Integrated Localization Environment
rst version had a server component that enabled the creation
of server projects. Products of Kilgray included MemoQ, MemoQ server,
QTerm, and translation memory Repository (
The development of translation technology
nding meaning-based translated material for reuse. LingotekÕs language search
engine indexed linguistic knowledge from a growing repository of multilingual
its database of previously translated material to Þ nd more speciÞ
c combinations
of words to reuse. Such meaning-based searching helped the translators to
tasks. Across Systems GmbH and MadCap Software announced a partnership to
combine technical content creation with advanced translation and localization.
Shortly after, Alchemy Software Development Ltd. and MadCap Software Inc.
announced a joint technology partnership that combined technical content cre-
ation with visual translation memory technology.
In 2008, Europe again Þ gured prominently in computer-aided translation
software production. In Germany, Across Language Server 4.0 Service Pack 1
was released in April, comprising, in addition to authoring, a number of exten-
sions, such as FrameMaker 8 and SGML support, context matching, and improve-
ments for web-based translations via crossWeb (MultiLingual 2008a). Three
months later, Across also introduced its new Language Portal Solution, which
would later be known as Across Language Portal, aimed at large-scale organiza-
at an international scale and implement Web portals for all language-related
issues and for all the members of the staff who needed to make use of language
resources at any level. In Luxembourg, Wordbee S. A. was founded as a transla-
tion software company focusing on web-based integrated computer-aided trans-
lation and management solutions ( In Ireland,
Alchemy Software Development, a company in visual localization solutions,
released in July Alchemy Publisher 2.0, which combined visual localization
technology with translation memory for documentation. It supported standard
documentation formats, such as MS Word, XML, application platforms such as
Windows 16/22/64x binaries, web-content formats such as HTML, ASP, and
all derivative content formats (
During this year, Eastern Europe made considerable progress in computer-
aided translation development and commercialization. In March, the Russian
PROMT released the version 8.0 with major improvements in the translation
engine, an enhanced translation memory system with TMX Þ les import support,
and extended language variant support, being able to deal with English (British
and American), Spanish (Castilian and Latin American), Portuguese (Portuguese
and Brazilian), German (German and Swiss), and French (French, Swiss, Belgian,
Canadian) documents ( Kilgray Translation Technolo-
gies, from Hungary, released in September MemoQ 3.0, which included a new
termbase and provided new terminology features. It introduced full support for
XLIFF as a bilingual format and offered the visual localization of RESX Þ
MemoQ 3.0 was available in English, German, Japanese, and Hungarian (http:// In Ukraine, Advanced International Translations (AIT) released
in December AnyMem, a translation memory system compatible with Microsoft
In Asia, Yaxin CAT 4.0 was released in China in August with some new
features, including a computer-aided project platform for project management
and huge databases for handling large translation projects. In Taiwan, Otek
released Transwhiz 10 for translating English, Chinese, and Japanese languages,
with a fuzzy search engine and a Microsoft Word workstation (http://www.
The development of translation technology
In North America, JiveFusion Technologies, Inc. in Canada ofÞ
cially launched
2007 Suite. New features included Context Match, AutoSuggest, and QuickPlace
( In addition, SDL released in October its enterprise
platform SDL translation memory Server 2009, a new solution to centralize,
share, and control translation memories (
In Canada, JiveFusion Technologies Inc. released Fusion 3.1 to enhance its
TMX compatibility and the capability to import and export to TMX while
The development of translation technology
included the DeepMiner data extraction engine, a new StartView interface, and
AutoWrite word prediction. In Hungary, Kilgray Translation Technologies
released in February MemoQ 4.0, which was integrated with project manage-
ment functions for project managers who wanted to have more control and
enable translators to work in any translation tool; and later that year, the company
released MemoQ 4.5, which had a rewritten translation memory engine and
improvements to the alignment algorithm ( In Swit-
zerland, STAR Group released, also in October, Transit NXT Service Pack 3 and
TermStar NXT. Transit NXT Service Pack 3 contained the following improve-
ments: support of Microsoft OfÞ
ce 2007, InDesign CS5, QuarkXpress 8, and
QuarkXpress 8.1, as well as PDF synchronization for MS Word Þ
les. In the
United Kingdom, SDL released in March a new subscription level of its SDL-
Trados Studio, which included additional productivity tools for translators such
ervice Pack 2, enabling translators to plug in to multiple automatic translation
In South America, Maxprograms in Uruguay released in April SwordÞ
sh II,
which incorporated Anchovy version 1.0Ð0 as a glossary manager and term
extraction tool, and added support for SLD XLIFF and Microsoft Visio XML
Drawings Þ les, amongst others (
The year 2011 witnessed computer-aided translation advances, particularly
in Europe. In Luxembourg, the Directorate-General for Translation of the
European Commission released in January its one million segments of multi-
lingual translation memory in TMX format in 231 language pairs. Translation
units were extracted from one of its large shared translation memories in
Euramis (European Advanced Multilingual Information System). This database
Acquis Communautaire
the entire body of European legislation, plus some other documents that were
not part of the
. Merely one month later, in Switzerland, the STAR
Group released the Service Pack 4 for Transit NXT along with TermStar NXT.
Transit NXT Service Pack 4 contained the following improvements: support
of MS OfÞ ce 2010, Support of Quicksilver 3.5l, and Preview for MS OfÞ
formats. In March, XTM 5.5 was released in the United Kingdom, in both
Cloud and On-Premise versions, which contained customizable workß
ows, a
new search and replace feature in Translation Memory Manager, and the
redesign of XTM Workbench ( In France, Atril/
The development of translation technology
database, a reinforced fuzzy match window, and adjusted buttons (http://
for Microsoft Word, was now replaced by the new Memsource Editor, the
abovementioned standalone translation editor. Further new improvements offer
the opportunity of adding comments to segments, along with an enhanced
version control, translation workß
The development of translation technology
The systematic compatibility with Windows and
Microsoft OfÞ
The development of translation technology
Brace, Colin (1993) ÔTM/2: Tips of the IcebergÕ,
Language Industry Monitor
MayÐJun, Available from
Brace, Colin (1994) ÔBonjour, Eurolang OptimizerÕ,
Language Industry Monitor

Issue MarÐApr, Available from
Bruderer, Herbert E. (1975) ÔThe Present State of Machine Translation and Machine-
aided TranslationÕ,
The development of translation technology
LISA (2010) ÔIBM and the Localization Industry Standards Association Partner to
Deliver Open-Source Enterprise-Level Translation ToolsÕ, Available from http://
Locke, William Nash and Andrew Donald Booth (eds.) (1955)
Machine Translation
of Languages: Fourteen Essays
, Cambridge, MA: MIT Press.
uk, Jurij N. (1989) ÔMachine-aided Translation: A Survey of Current SystemsÕ,
Istvan S. Batori, Winfried Lenders, and Wolfgang Putschke (eds.)
Linguistics: An International Handbook on Computer-oriented Language Research
, Berlin and New York: De Gruyter, 682Ð688.
Melby, Alan K. (1978) ÔDesign and Implementation of a Machine-assisted Transla-
Seventh International Conference on Computational
held in Bergen, Norway, 14Ð18 August.
Melby, Alan K. and Terry C. Warner (1995)
of the Nature of Language, with Implications for Human and Machine Translation
Amsterdam and Philadelphia: John Benjamins Publishing Company.
MultiCorpora Inc. (2011) ÔMultiCorpora Launches New Translation Management
SystemÕ, Available from MultiCorporaÕs website at http://www.multicorpora.
When the term Ôcomputer-aided translationÕ is mentioned, we often associate
it with the functions that a computer-aided translation system can offer, such
dictionaries and browsers; and the computational hitches that we often encounter
when working on a computer-aided translation project, such as chaotic codes.
However, what is more important is to see beyond the surface of computer-
functions in translation technology.
Concepts, which are relatively stable, govern or affect the way functions are
designed and developed, and functions, which are fast-changing, realize, in turn,
the concepts through the tasks they perform. As a major goal of machine trans-
lation is to help human translators, a number of functions in computer-aided
translation systems have been created to enable machine processing of the source
with minimal human intervention. Concepts, moreover, are related to what
translators want to achieve in translating. We have identiÞ
ed seven major con-
cepts which are of particular importance in computer-aided translation: simula-
tivity, emulativity, productivity, compatibility, controllability, customizability, and
collaborativity. These concepts are arranged in this order for easier memorization
by their acronym, SEPCCCC. Simply put, translators want to have a control-
with Þ le formats and language requirements (
part of the human translator and the creation of a number of quality assurance
tools to follow the way checking is performed by a human translator.
There are a number of ways to illustrate manÐmachine simulativity.
Goal of translation
The Þ rst is about the ultimate goal of translation technology. All forms of
translation, including machine translation, computer-aided translation, and
machine translation, the goal of a fully automatic high-quality translation
human intervention. In the case of computer-aided translation, the same goal
is to be achieved with a computer-aided translation system that simulates the
behaviour of a human translator through manÐmachine interaction.
Translation procedure
A comparison of the procedures of human translation with those of computer-
aided translation shows that the latter simulates the former in a number of ways.
In manual translation, various translation procedures have been proposed by
translation scholars and practitioners, ranging from two-stage to eight-stage
machine translation and computer-aided translation, the process is known as
technology-oriented translation procedure.
Two-stage model
In human translation, the Þ rst type of translation procedure involves a two-stage
model, which consists of a stage of source text comprehension and a stage of
Major concepts in computer-aided translation
The Þ rst model of a three-stage translation procedure, involving the three phases
of analysis, transfer, and restructuring, was proposed by Eugene Nida and Charles
Taber (1969/1982: 104). They intended to apply elements of ChomskyÕs
transformational grammar to provide Bible translators with some guidelines
Major concepts in computer-aided translation
data creation, and the creation of terminology and translation memory databases.
At the third stage, the data processing stage, the tasks include data analysis,
using system and non-system dictionaries, using concordancers, pre-translating,
data processing by computer-aided translation systems with human intervention
or by machine translation systems without human intervention, or data process-
ing by localization systems. At the fourth stage, the data editing stage, the work
is divided into two types. One type is data editing for computer-aided translation
systems, which concerns interactive editing and the editing environments, match-
are created by IMG tag (inline image graphic tag), and the text that provides
an alternative message to viewers who cannot see the graphics is known as ALT
tag, which stands for Ôalternative textÕ. Adding an appropriate ALT tag to every
image within oneÕs website will make a huge difference to its accessibility. As
translators, our concern is the translation of the alternative text, as images are
not to be translated anyway.
Chatroom translation
Machine translation has the function to translate the contents of a chatroom,
known as Ôchat translationÕ or Ôchatroom translationÕ. Chat translation sys-
tems are commercially available for the translation of the contents of the
chatroom on the computer. As a chat is part of conversational discourse,
Major concepts in computer-aided translation
that can be handled by a system can be relatively large. Fluency, for example, sup-
ports the conversion of currencies of around 220 countries.
Email translation
Email translation refers speciÞ cally to the translation of emails by an MT system
rst online and real-
time email translation was achieved in 1994 by the CompuServe service, which
Graehl 1997, 1998; Kuo, Li, and Yang 2006; Li, Zhang, and Su 2004: 159Ð166;
Lin and Chen 2002; Lin, Wu, and Chang 2004: 177Ð186). In the case of Chi-
ed characters
is used mostly in mainland China, whereas the Wade-Giles Romanization system
for traditional characters is used in Taiwan (Lee and Chang 2003: 96Ð103).
Mouse translation
This feature involves translating sentences on a webpage or application by simply
clicking the mouse. Atlas is an example of systems that provide mouse translation.
Online translation
This is the translation of a text by an online MT system that is available at all
times on demand from users. With the use of online translation services, the func-
tions of information assimilation, message dissemination, language communication,
translation entertainment, and language learning can be achieved (Clements 1996:
220Ð221; Flanagan 1995, 1996: 192Ð197, 1997; Gaspari 2004: 68Ð74; McCarthy
2004; Mellebeek, Khasin, van Genabith, and Way 2005; OÕNeill-Brown 1996:
222Ð223; Zervaki 2002).
Machine translation is regarded as pre-translation on two counts. The former
involves preparatory work on the texts to be translated, including checking the
spelling, compiling dictionaries, and adjusting the text format. The latter is
taken to be a draft translation of the source text, which can be further revised
by a human translator.
Sentence translation
performs sentential translation. In other words, machine translation works on
Major concepts in computer-aided translation
communication, to enable people with different language backgrounds to
communicate with each other.
As translation technology is a Þ eld of entrepreneurial humanities, productivity
is of great importance. Productivity in computer-aided translation, in particular,
is achieved through the use of technology, collective translation, recycling trans-
among translators. As this is the case, translators do not have to produce their
own translations. They can simply draw from and make use of the translations
stored in the bilingual database to form their translation of the source text.
Translation is therefore produced by selection.
Reusing translations to increase productivity
To reuse a translation in computer-aided translation is to appropriate terms
and expressions stored in the term database and translation memory database
(Craig, Dorr, Lin, Pavel, and Hajic 2006). It should be noted that, in literary
translation, translators produce translations in a creative manner. In practical
translation, however, translators reuse and recycle previously translated seg-
Major concepts in computer-aided translation
and the total number of computer-aided translators in the world is likely to be
Seeking proÞ t to increase productivity
Translation is in part vocational, in part academic. In the training of translators,
there are courses on translation skills to foster their professionalism, and there
are courses on translation theories to enhance their academic knowledge. How-
ever, there are very few courses on translation as a business or as an industry. It
should be noted that translation in recent decades has increasingly become a
eld of entrepreneurial humanities as a result of the creation of the project
management function in computer-aided translation systems. This means that
translation is now a Þ eld of humanities which is entrepreneurial in nature. Trans-
lation as a commercial activity has to increase productivity to increase proÞ
Saving labour to increase productivity
Computer-aided translation systems help to increase productivity and proÞ
The concept of compatibility in translation technology must be considered in
terms of Þ le formats, operating systems, translation memory databases, terminol-
ogy databases, and languages supported by different systems.
Compatibility of Þ
le formats
One of the most important concepts in translation technology is the type of
data that needs to be processed, which is indicated by its format, being shown
Major concepts in computer-aided translation
Some computer-aided translation systems that can handle PowerPoint Þ
les are
Adobe InDesign is a desktop publishing program. Alchemy Publisher and Any-
Mem are able to translate the documents created by InDesign without the need
of any third party software. However, for Alchemy Publisher, .
les must
be exported to an .
format before they can be processed. Other computer-
aided translation systems that support Adobe InDesign Þ
les include Across, DŽjˆ
Vu, Fortis, GlobalSight, Heartsome Translation Suite, MemoQ, MultiTrans,
Okapi Framework, SDL-Trados, SwordÞ sh, Transit, and XTM.
Major concepts in computer-aided translation
Adobe FrameMaker is an authoring and publishing solution for XML. FrameMaker
les, characterized by the extensions .
, can be opened directly
by a translation system if Adobe FrameMaker has been previously installed.
Some examples of computer-aided translation systems that can translate Adobe
Software development types
Java Properties Þ les are simple text Þ les that are used in Java applications. The
le extension of Java Properties Files is .
DŽjˆ Vu, Fortis, Heartsome Translation Suite, Lingotek, Okapi Framework,
OmegaT+, Open Language Tools, Pootle, SwordÞ
sh, and XTM support Java
Properties Þ
ce, of the Star Division, was a German company that ran from 1984 to, an open-source version of StarOfÞ
owned by Sun Microsystems, from 1999 to 2009, and by Oracle Corporation
from 2010 to 2011. Currently it is known as Apache OpenOfÞ ce. The exten-
sion of the OpenOfÞ ce format is .
(OpenDocument Format).
Some computer-aided translation systems that are able to process this type
Major concepts in computer-aided translation
in 1992, is a typical example of a computer-aided translation system working
DOS was supplemented by Microsoft Windows 1.0, a 16-bit graphical operat-
ing environment, released on 20 November 1985 (Windows 2012). In November
1987, Windows 1.0 was succeeded by Windows 2.0, which was available until
2001. DŽjˆ Vu 1.0, released in 1993, was one of the Þ
rst systems compatible
with Windows 2.0. Windows 2.0 was later supplemented by Windows 286 and
Windows 386.
Then, on 22 May 1990, came Windows 3.0, succeeding Windows 2.1x.
Windows 3.0, featuring a graphical environment, was the third major release of
Microsoft Windows. With a signiÞ cantly revamped user interface and other
technical improvements, Windows 3 became the Þ
rst widely successful version
of Windows and a rival to AppleÕs Macintosh and the Commodore Amiga on
the GUI front. It was followed by Windows 3.1x. During its lifespan from 1992
to 2001, Windows 3.1x introduced various enhancements to the still MS-DOS-
based platform, including improved system stability, expanded support for mul-
IBM operating system
OS/2 is a series of computer operating systems, initially created by Microsoft
and IBM, then later developed by IBM exclusively. Its name stands for ÔOperat-
Until 1992, the early computer-aided translation systems ran either on MS-
DOS or OS/2. For example, IBM Translation Manager/2 (TM/2) was released
Major concepts in computer-aided translation
Compatibility of terminology databases
Compatibility of terminology databases is best illustrated by TermBase eXchange
(TBX), which covers a family of formats for representing the information in a
high-end termbase in a neutral intermediate format, complying with the Ter-
TBX is an international standard as well as an industry standard. The industry
standard version differs from the ISO standard only by having different title pages.
LISA, the host organization for OSCAR that developed Termbase Exchange, was
dissolved in February 2011. In September 2011, the European Telecommunications
a large number of languages and sub-languages in the world, totalling 6,912.
But the number of major languages that computers can process is relatively small.
Major concepts in computer-aided translation
Percentage of
Percentage of Non-
Across1211861 (50.41%)60 (49.59%)
DŽjˆ Vu1322166 (50%)66 (50%)
54 (52.94%)48 (47.06%)
48 (53.33%)42 (46.67%)
62 (53.91%)53 (46.09%)
54 (59.34%)37 (40.66%)
68 (43.31%)89 (56.69%)
Major concepts in computer-aided translation
Controlled language is used by both humans and computers. The goals of using
controlled language are to make the source text easier to read and understand.
These goals are to be achieved both at the lexical and sentential levels.
At the lexical level, controlled language is about the removal of lexical ambiguity
and the reduction in homonymy, synonymy, and complexity. This is to be achieved
by one-to-one correspondence in the use and translation of words, known as one-
word one-meaning. An example is to use only the word ÔstartÕ but not similar
Major concepts in computer-aided translation
Engines Ltd., the Controlled English of Alcatel Telecom, and the Boeing SimpliÞ
English Checker of the Boeing Company (Wojcik and Holmback 1996: 22Ð31).
For commercial controlled language checkers, there are a number of popular systems.
The LANTmaster Controlled Checker, for example, is a controlled language checker
Controlled language in computer-aided translation systems
The concept of controlled language is developed by controlled authoring tools
in computer-aided translation systems. Authoring tools are used to check and
improve the quality of the source text. There is an automatic rewriting system
which is usually used as a tool to realize controlled authoring. One of the
computer-aided translation systems that performs controlled authoring is Star
Transit. This open-source system provides automatic translation suggestions
from the translation memory database from a speedy search engine and can
Major concepts in computer-aided translation
will have implications on resources, such as time and labour. Localization is also
an important factor to be considered. Very often, what is stored in a database
prepared by translators of a speciÞ c region may not be usable in another region.
In this case, efforts are needed to prepare new databases to deal with regional
variations in terminology and translation memory.
Language customization
It is true that there are many language combinations in computer-aided translation
a customized dictionary is an enormous task, involving a huge amount of work
in database creation, maintenance, and database updating.
Syntactical customization
Syntactical customization, on the other hand, involves adding sentences or phrases
Major concepts in computer-aided translation
with higher efÞ ciency. The best example to illustrate this point is project col-
laboration, which allows translators and project managers to easily access and
distribute projects, as well as monitor easily the progress of these projects.
The translation work in the present digital era is done almost entirely online
with the help of a machine translation or computer-aided translation systems.
This can be illustrated with SDL-Trados 2007 Synergy, which is a computer-
aided translation system developed by SDL International and generally considered

Figure 2.1
Screenshot of Dashboard of SDL-Trados 2015

Figure 2.2
Screenshot of SDL-Trados: ÔProgress of Individual ProjectsÕ

Figure 2.3

Figure 2.4
Screenshot of SDL-Trados: ÔProgress on the Translation of FilesÕ
Workß ow of a translation project
To start a project, the Þ rst stage of the workß ow is the creation of a termbase
and a translation memory database. In other words, when the project manager
has any publications, Þ les, or webpages to translate, he or she will send them
either to the in-house translators in a department or unit, or to freelancers for
processing. They will create translation units and term databases from these
pre-translated documents and save these databases in the SDL-Trados 2007
Server. This is the Þ rst stage of the workß
Major concepts in computer-aided translation
After the creation of a translation memory and a termbase, the project man-
ager can then initiate the translation project and monitor its progress with the
use of SDL-Trados 2007 Synergy. He or she can assign and distribute source
les to in-house or freelance translators by email. Translators can then perform
the translation by (i) reusing the translation memories and terms stored in the
databases, and (ii) adding new words or expressions to the translation memory
Adriaens, Geert and Lieve Macken (1995) ÔTechnological Evaluation of a Controlled
Language Application: Precision, Recall and Convergence Tests for SECCÕ,
Major concepts in computer-aided translation
Carbonell, Jaime G., Teruko Mitamura, and Eric H. Nyberg (1992) ÔThe KANT Per-
spective: A Critique of Pure Transfer (and Pure Interlingua, Pure Statistics, . . . )Õ,
), Montreal, Quebec, Canada,
Flanagan, Mary A. (1997) ÔOnline Translation: MTÕs New FrontierÕ,
, London: The Association for Information Management.
Fouvry, Frederik and Lorna Balkan (1996) ÔTest Suites for Controlled Language
Proceedings of the 1st International Workshop on Controlled Language
Gaspari, Federico (2004) ÔEnhancing Free On-line Machine Translation ServicesÕ,
Proceedings of the 7th Annual CLUK Research Colloquium
of Birmingham, Birmingham, the United Kingdom, 68Ð74.
Gommlich, Klaus (1989) ÔThe Control Potential of a Computer-aided Translation
Goto, Isao, Naoto Kato, Noriyoshi Uratani, and Terumasa Ehara (2003) ÔTrans-
literation Considering Context Information Based on the Maximum Entropy
Major concepts in computer-aided translation
Workshop on
), Language Technolo-
gies Institute, Carnegie Mellon University, Pittsburgh, PA, the United States of
America, 51Ð61.
Kay, Martin (1980) ÔThe Proper Place of Men and Machines in Language Transla-
Research Report CSL-80-11
, Xerox Palo Alto Research Center, Palo Alto,
Kertesz, Francois (1974) ÔHow to Cope with the Foreign-language Problems:
Experience Gained at a Multidisciplinary LaboratoryÕ,
Journal of the American
Lux, Veronika and Eva Dauphin (1996) ÔCorpus Studies: A Contribution to the
nition of a Controlled LanguageÕ,
Proceedings of the 1st International Workshop
Matsuda, Junichi and Hiroyuki Kumai (1999) ÔTransfer-based Japanese-Chinese
Translation Implemented on an E-mail SystemÕ,
MT in the Great Translation Era
, Singapore.
McCarthy, Brian (2004) ÔDoes Online Machine Translation Spell the End of Take-
home Translation Assignments?Õ
Melby, Alan K. (1995)
The Possibility of Language: A Discussion of the Nature of
Language, with Implications for Human and Machine Translation
, Amsterdam
and Philadelphia: John Benjamins Publishing Company.
Melby, Alan K. (2012) ÔTerminology in the Age of Multilingual CorporaÕ,
Journal of Specialized Translation
Mellebeek, Bart, Anna Khasin, Josef Van Genabith, and Andy Way (2005) ÔImprov-
ing Online Machine Translation SystemsÕ,
Major concepts in computer-aided translation
Nyberg, Eric H., Teruko Mitamura, and Jaime G. Carbonell (1997) ÔThe KANT Machine
System: From R&D to Initial DeploymentÕ,
LISA Workshop on Intergrating Advanced
Translation Technology
, Seattle, Washington, DC, the United States of America.
Nyberg, Eric H., Teruko Mitamura, and Willem-Olaf Huijsen (2003) ÔControlled
Language for Authoring and TranslationÕ, Harold L. Somers (ed.)
Translation: A TranslatorÕs Guide
, Amsterdam and Philadelphia: John Benjamins
Publishing Company, 245Ð281.
OÕNeill-Brown, Patricia (1996) ÔOnline Machine TranslationÕ,
Conference of the Association for Machine Translation in the Americas: Expanding
MT Horizons (AMTA-2)
, Montreal, Quebec, Canada, 222Ð223.
Probst, Katharina and Lori S. Levin (2002) ÔChallenges in Automated Elicitation
of a Controlled Bilingual CorpusÕ,
Proceedings of the 9th International Conference
Wilss, Wolfram (1982)
The functions in a computer-aided translation tool are instructions or com-
mands given to the system so that it performs the actions the user wants it to
perform. Functions are developed according to some of the concepts mentioned
in the previous chapter. However, concepts or goals in translation technology
cannot be achieved without the support of an adequate amount of data. All
actions in translation technology are data-based and data-driven. It is, therefore,
logical to divide the entire process of translation in terms of data into an ini-
tiating stage, a data preparation stage, a data processing stage, a data editing
nalizing stage, putting the functions under the stage to which
they belong. This function-based approach is ideal to understand which functions
It allows us to identify functions for starting up a
cations and creating new projects, or opening old
projects, during the Þ
rst stage; functions for data collection and data creation,
during the second stage; functions for data analysis, data formats, data mining,
data reuse, and data application, in the third stage; functions for data editing,
in the fourth stage; and functions for data updating and data delivery, in the
nalizing stage.
The following discussion about the functions of computer-aided translation
systems will be based on the nine most popular translation software on the
Initiating stage
ve stages in the process of computer-aided
translation. The functions of this stage are wizards for the creation of projects,
translation memory databases, and terminology databases.
able to use their service. Licensing, as in the case of XTM (User Manual 2011:
13), depends on the type of account purchased. Named users, for instance, are
freelancers and small-group accounts. The number of users is speciÞ ed in the
subscription agreement. Concurrent users are for language-service-providers
For cloud-based systems, the user needs a web address to access the system
Functions in computer-aided translation systems
rm the registration of the userÕs
computer. Then the user will be redirected to a page to change his password.
A number of measures have been designed to control access to the system.

Name of the SystemName of the Project WizardFile Extension of the Project
AcrossProject WizardNil (project name is
shown in crossBoard)
DŽjˆ Vu X2New Project Wizard
New memoQ Project Wizard
OmegaTCreate New Project
SDL-Trados Studio 2014Project Creation Wizard
Snowman 1.3New Project
Wordfast Classic 6.03tCreate Project
Yaxin CAT 4.0
Create Project
Functions in computer-aided translation systems
Functions in computer-aided translation systems
Functions in computer-aided translation systems
due to segmentation (deciding where one word ends and the next begins)
and variability (variations in how an individual word can be pronounced Ð loud
or soft, fast or slow, with rising or falling intonation).
Importing source text/document
Another way of text creation is to import source document or source text into
the source text pane. There are two ways to do text importation: loading it from
le menu or copying and pasting it. In the former, a source text, provided
that it is part of a Þ le, can be selected, loaded, and imported into the source
pane. In the latter, the text can be copied from one location and pasted onto
the source text pane.
Creation of terminology databases
In computer-aided translation, terminology refers to the terms in the source
language created from project-based documents (Ar‡ujo 2000; Bowker 2006a,
2006b; Gillam, Ahmad, Dalby, and Coz 2002; Maia 2003: 43Ð53; Pearson
1998b: 258Ð262; Zauberga 2005: 107Ð116). In Across, it is deÞ
ned as Ôthe
total stock of concepts and the respective terms in a subject areaÕ (Manual 2012:
Terminology also refers to the study of the body of specialized words relating
to a particular subject, which are processed electronically to be used in computer-
There are several ways to create a terminology database. For instance, a
computer-aided translation system can analyse a project and automatically extract
words and short phrases from it. By displaying them in order of frequency,
important terms and expressions are easily spotted, while non-relevant ones can
Term extraction tools
There are term extraction tools within the computer-aided translation systems
in the form of both term managers, as well independent term creation tools,
not linked to any particular computer-aided translation system (Bowker 2003:
49Ð65; Thurmair 2003). Terms can, thus, be extracted by the use of tools,
within or without translation software.
The terms extracted and stored form terminology databases, otherwise known
as Ôterm banksÕ or Ôterminology banksÕ. These databases contain translations of
single words or short phrases and are used to translate all the input terms and
improve the quality of fuzzy matches. The way terms in a translation system
are handled is known as terminology management, which refers speciÞ
cally to
the documentation, storage, manipulation, and presentation of a specialized
terminology or glossary in a computer-aided translation system. There are ter-
minology management systems for the user to create, maintain, and search for
terms in bilingual or multilingual databases.
As aforementioned, there are also tools that can be used separately to create
terminology databases. One of these tools, WordSmith, enables the user to Þ
out how words are used in certain texts. This application offers a ÔWordlistÕ
Functions in computer-aided translation systems
In order to add a term to a terminology database while working within an
editing environment, the user must select the word and choose the function
key ÔAdd TermÕ. The chosen word will then appear in the ÔSourceÕ Þ
eld. To
enter a translation for the term, the user can either type the translation in the
Some systems, such as Across, can create a terminology report.
The Search bar offers various options for Þ nding entries and terms.
Functions in computer-aided translation systems
Features of terminology databases
By making use of the materials stored in terminology databases, translators can
t from computer-aided translation tools in the following:
(1) Terminological consistency, which means that the same term is always used
the same term.
(2) Access rights to terminology databases can be limited. Customer-speciÞ
terminologists can only access the terminology of one customer, while global
experts can access all the terms in the system for all customers. The access
rights include Ôadd terminologyÕ, Ôexport terminologyÕ, Ôimport terminologyÕ,
Ômodify terminologyÕ, and Ôview terminologyÕ.
(3) Some systems allow the use of multiple terminology databases. Wordfast
Classic, for example, has three simultaneous glossaries.
Storage of terminology databases
Different systems use different names to designate their terminology databases.
Table 3.2 shows the Þ les extensions of the eleven selected systems.

Size of terminology databases
The maximum size of a terminology database varies from system to system. For
Wordfast, it is voluntarily limited to 250,000 entries. For most systems, the size
of terminology databases is close to unlimited.

Table 3.2
Designations of Terminology Databases of Nine CAT Systems
Name of the Terminology
the Terminology
Across 5.5
DŽjˆ Vu X2
Fluency Translation
Personal Terminology
Term Bases
SDL-Trados Studio 2011SDL-Trados MultiTerm
Project Dictionary
WordfastClassic 6.0
Yaxin CAT 4.0
Creating translation memory databases
The process of creating a translation memory database goes through the stages
of segmentation and alignment, involving the use of tools, such as concordanc-
Segmentation consists of sentence separation in machine translation or computer-
ned unit of a source text that can be aligned with its corresponding
translation in a machine translation system or a computer-aided translation tool.
This is a function that splits the source text into segments using the rules and
Functions in computer-aided translation systems
502Ð509; Kashioka 2005; Lin and Cherry 2003; LŸ, Zhao, Li, and Yang 2001:
108Ð115; Sun, Jin, Du, and Sun 2000: 110Ð116; Taskar, Simon, and Dan 2005;
Wu and Wang 2004: 262Ð271). Phrase alignment, as indicated in the name
itself, refers to the alignment of a phrase in the source text with its equivalent
Some important concepts about translation memories will be provided
A translation memory database, which is created either as a project Þ
le to be
used in a speciÞ c project or as a standalone Þ le to be used with a speciÞ
c project
as well as other related projects, has the following features. First, it is a reusable
to be translated, the system scans the database for a previous source segment
that matches the new segment exactly (exact match) or approximately (fuzzy
match). Third, it contains possible usable translations from exact and fuzzy
matches. Fourth, the translator has the choice of accepting, modifying, or reject-
ing the suggested translations. Fifth, it is a sentence storehouse, as it is a database
that stores sentences already translated in a translation memory system. Sixth, it
Functions in computer-aided translation systems
Different designations for translation memory databases have been used by dif-
ferent computer-aided translation systems in data storage, such as
Across, which is deÞ ned as a central databank for the storage of translated sentences

Table 3.3 provides a list of names for translation memory databases in different

The size of translation memory databases varies from system to system. Wordfast
Classic, for example, can store up to 1,000,000 units per single translation
memory. The user can create and maintain as many translation memories as the
There are different types of translation memory, some of which are introduced

Table 3.3
Designations of Translation Memory Databases of Nine CAT Systems
Name of the Translation
Memory Function
Translation Memory
Across 5.5crossTankimport:
DŽjˆ Vu X2Translation Memory*.
MemoQ 6.2Translation MemoriesImport: .
Export: csv
SDL-Trados Studio 2013Translation Memories*.
Import: .
sdlxliff, .ttx
Export: .
(Project Translation
Memory Database)
Import: .
Wordfast Classic 6.0
Translation MemoryNil
Yaxin CAT 4.0
(1) Background translation memory
This is one of the three types of translation
memories offered by Wordfast; the other two being Translation Memory (TM)
and Very Large Translation Memory (VLTM).
A background translation memory is a read-only translation memory, which
Wordfast will scan for an exact match before scanning the translation memory
in use. If both the background translation memory and the current translation
memory yield translation units with the same match rate, the formerÕs transla-
tion units will be displayed. It is worth noting, however, that WordfastÕs user
can specify or change the order of preference for the three types of translation
memories within the system (User Manual of Wordfast Classic Version 6.0 2013:
29). A background translation memory can be created by selecting a translation
memory database other than the active translation memory. It provides exact
(2) File-based translation memory

This refers to a type of translation memory
which is based on a group of Þ les in a computer-aided translation system.
(3) Master translation memory

This refers to a big translation memory ranging
from 100,000 to 1,000,000 translation units.
(4) Server-based translation memory

This refers to a type of translation
memory that is based on a group of databased tables on a database server in a
Functions in computer-aided translation systems
(6) Working translation memory

This refers to a small translation memory
which has all the translation units close to the text. To prepare this database,
a computer-aided translation system segments the text into translation units
It goes without saying that analysis is important in both human and machine
translation. The use of concordancers has facilitated greatly the work of analysis.
Generally and computationally, it is necessary to conduct source analysis (Bernth
and McCord 2000: 89Ð99; Mitamura, Baker, Svoboda, and Nyberg 2003;
Nyberg, Mitamura, Svoboda, Ko, Baker, and Micher 2003), text analysis (Hou
M. 2001: 204Ð210; Taylor and Baldry 2001: 277Ð305; Wonsever, and Minel
2001; Yarowsky, Ngai, and Wicentowski 2001: 109Ð116), natural language
Functions in computer-aided translation systems
equivalents (Varantola 1998: 179Ð192). The same holds true for computer-aided
translation (Gliozzo and Strapparava 2006).
Before we discuss the role of dictionaries in translation technology, an
explanation of the differences among dictionaries, glossaries, and terminologies
may be in order. According to the
WebsterÕs Third New International Diction-
HanYing fenlei Chatu cidian
ed and Illus-
trated Chinese-English Dictionary
) (Compiling Group 1981), the reverse dic-
tionary of
A Reverse Chinese-English
) (Yu Y. 1986), and the visual dictionary of the
Visual Dictionary
(Corbeil and Archambault 2006). Another categorization
Functions in computer-aided translation systems

: A headword/character is a word or term, often in a distinctive
type, placed at the beginning of an entry. The main function of a headword
nd the word or term they want.

word to indicate homographs. It tells the user to look for other identical

: Illustrations, in the forms of pictures, diagrams, and graphs,
are meant to facilitate the comprehension of certain terms through
visual aids.

ected form
: Inß ected forms give other grammatical forms of the same
headword, indicating how to use the word correctly.

: A linguistic label is a code that refers to the style, register,

Part-of-speech marker
: This marker indicates the part of speech of the headword
with a code.

Functions in computer-aided translation systems
Ôreferential meaningÕ. It is, nevertheless, difÞ
cult to establish what cognitive
meaning is and deÞ ne it by referring it to physical properties. Most adjec-
tives and adverbs do not have cognitive meanings.

consists of the associations that a word acquires on
account of the meanings of the words which tend to occur.

Combinatory meaning
is a type of lexical meaning that results from the com-
bination of words.

refers to the associations related to a word.

is the meaning that is derived from the context, such as
the meaning of a word within a particular sentence, or the meaning of a
sentence within a particular paragraph. Contextual meaning is of great
importance for translators.

Core meaning
is the most basic meaning of the word.

concerns the situation in which words that originally have deroga-
tory or neutral meanings develop favourable meanings in the course
of time.

refers to the charge of feeling carried by a particular word
or expression in a given utterance or text (Nida 1964: 70Ð119).

concerns words that bear implications of either approval
or disapproval.

concerns things that a word can be extended to apply to. For

is said to be present if a word in one language does not seem to
have a counterpart in another language. Lexical gaps can usually be Þ
by paraphrases.

refers to the surface meaning of a word or an expression.

Narrowing of meaning
concerns the narrowing of the denotational scope of a
lexeme. This often takes place when part of the denotational domain of a
lexeme is being taken up or usurped by the extension of the scope of another
lexeme, resulting in lexical specialization.

Primary meaning
is the meaning of a word at its creation, or the meaning
that usually comes to mind when the word is said in isolation.

Functions in computer-aided translation systems
to K. Balasubramanian, ÔAbsolute equivalence requires that the lexical
unit be identical in all the three components of lexical meaning, viz.,
designation, connotation and range of application and occur in all the
typical contexts in which SL lexical unit occursÕ (Balasubramanian 1988:
13Ð20). Some examples of absolute equivalents are as follows: numer-
; ordinals, such as ÔÞ
; parts of the body, such as ÔheadÕ
, and ÔheartÕ
; measurements, such as ÔyardÕ
of prominent politicians are usually translated by government translators
cial documents.

concerns the use of equivalents that involve the pro-
Functions in computer-aided translation systems
found in the above dictionaries. Atlas allows the user to create up to 1,000
c translation
projects (
Some computer-aided translation tools allow the user to deÞ
ne user-speciÞ c
translations; study language usage in translation; analyse the linguistic features
in translation through the quantiÞ cation of lexical and syntactical features;
Functions in computer-aided translation systems
usage; (2) it helps to prepare a translation memory dictionary; (3) it helps to
prepare a project-speciÞ c bilingual dictionary; and (4) it also deÞ nes the scope
of the vocabulary (Langlois 1996: 34Ð42).
Types of data processing
With the increasing number of ready-to-use dictionaries and concordancers
available to translators, data processing is now possible. There are three major
types of data processing:
, which goes through the stages of pre-translation and transla-
tion. Computer-aided translation systems are used for pre-translation and
translation and machine translation systems are for automatic translation.
Chapter 4 in conjunction with
computer-aided translation, in the form of hybrid systems, it will not be
discussed here.
to create localized software or webpages.
Multimodal Transfer
involves the use of systems to produce different forms
Functions in computer-aided translation systems
translatorÕ (Manual 2012: 91). ÔLeveraged MatchÕ is another term used in XTM
to refer to the situation in which Ôa sentence or phrase in a translation memory
is the same phrase in a different context as the sentence or phrase the translator is
currently working onÕ. This term is interchangeable with 100% leveraged matches.
Less than 100% matching is regarded as fuzzy
matching (Lin and Chen 2006). According to XTM, a fuzzy match Ôrefers to
the situation when a sentence or phrase in a translation memory is similar (but
not a 100% match) to the sentence or phrase the translator is currently working
onÕ (Manual 2012: 91). Fuzzy matches are further divided into three groups:
75Ð84%, 85Ð94%, and 95Ð99%, with the fuzzy matching starting at 74%.
Fuzzy matching in DŽjˆ Vu performs two functions. The Þ rst is ÔassembleÕ,
which involves examining the translation memory databases to take relevant
fragments or sentences with a similar structure to produce a translation from
which automatically examines the translation databases and takes relevant frag-
ments or sentences with a similar structure to produce a translation from material
Translation phase: Translating with computer-aided
Some issues involved in the use of computer-aided translation systems are
discussed below.
Functions in computer-aided translation systems
have to produce their own translations. They can simply draw from and
make use of an array of equivalents stored in a bilingual database to form
their translation of the source text. This has great signiÞ
cance to the pro-
fession and pedagogy of translation.

Consistency in Terminology and Style
can be achieved by translation teams
number of systems is relatively large. There are fourteen types of computer-aided
translation systems. Third, several changes have taken place in the industry.
Computer-aided translation systems are moving from sentence-based to text-
Functions in computer-aided translation systems
with ÔneverÕ, if ÔneverÕ was in the terminology database. This means that a fuzzy
match is turned into an exact match by ÔrepairingÕ it, i.e., substituting one word
a multilingual system that handles the major languages of the European community.
It combines all functionalities of the pre-translation server and translation workstation
on a single personal computer, and is fully compatible with OptimizerÕs client-server
version. It can also produce fuzzy matches, plus technical term assistance.

Memory-based machine translation systems
refer to machine translation systems
with the function of a translation memory. The Þ rst generation, including
Functions in computer-aided translation systems
are basically two types of memory-based machine translation system, classiÞ
according to the order in which translation memory is used:
(1) Automatic translation and translation memory-based machine translation
These systems translate by automatic translation before using the
Functions in computer-aided translation systems
(2) Web-based multilingual computer-aided translation systems
Yaxin can also
be used in a multilingual manner, with all the features and functions working very
The Þ rst observation we can make is that Chinese has become an important
language in computer-aided translation. According to an article in
Software localization concerns the process of adapting and translating a
software application into another language to make it linguistically and cultur-
Functions in computer-aided translation systems
adapt a product or programme to different countries and cultures or locales, thus
enabling a product at a technical level for localization (Kumhyr, Merrill, and Spalink
1994: 142Ð148). In practice, this means that, in order to make it neutral and
c information, such as currencies,
Software and web localization can be studied from Þ
ve aspects: (1) computational;
(2) linguistic; (3) cultural; (4) economic; and (5) legal.
Most programmers design software that addresses, generally, local requirements.
Software localization will certainly enlarge the pool of software users. The Þ
of software localization is growing fast. There are a large number of software
systems. Failing to localize software severely limits its sales potential. It is very
Functions in computer-aided translation systems

Codes can also pose problems in localization.
Characters in Indo-European languages are encoded by one byte; however,
in order to encode a single Chinese, Japanese, or Korean character, at
least two bytes are required. Chinese, for example, is double-byte, while
Japanese can be single-, double-, or even triple-byte for Kana and Kanji. As
a result, when English software is localized into Japanese, the translation
may be truncated as a result of buffer storage. This is due to the fact that, in
software localization, the buffer may not be able to accommodate more bytes
Graphics are symbols produced by a process, such as handwriting,
drawing, or printing. In the process of software localization, embedded graphics
are a frequent source of problems, and special consideration must be given to the
number and type of graphics. What is more, text should not be used alongside
graphics, such as the use of the word ÔhelpÕ when a question mark on the button
serves the same purpose.
An icon is a graphic presentation of an object on a computer
screen. Translating icons in HTML format is not an easy task for some
localization software.
(xiv) Hot-keys and short-cut keys
Adjustments may be required, as the names
of the keys are localized.
Some software may include audio, such as songs. When such
software is localized, the original sounds, dialogue, or songs might need to be

to insert the subtitles.
ÔLocalizationÕ, say Ashworth and OÕHagan (2002: 66), Ôcan be deÞ
ned as a
process to facilitate globalization by addressing linguistic and cultural barriers
c to the Receiver who does not share the same linguistic and cultural
background of the SenderÕ. Several aspects of linguistic localization are worth
Translating a text from one language into another
gives it a new life by overcoming linguistic barriers, revitalizing it on a different
soil. Localization is to give a product a life beyond a locality. In doing so, it is
important to identify the locales in which languages are used, such as the French
and Italian of the Western European scripts, Greek and Russian of the Eastern
European scripts, and Chinese and Japanese of the Asian scripts.
Functions in computer-aided translation systems
from left to right and right to left, and vertically from left to right or right to left.
When a text contains strings written from left to right, such as numbers or English,
it has to put them from left to right. The directionality of languages thus causes
typographical problems, which have to be resolved before the text can be localized.
Linguistic adjustments may be necessary due to
the limited screen space when localizing software. Some localization software,
(v) Typographical changes
Typographical or logographical changes may
be necessary when localizing from one language into another. The use of
abbreviations for different command buttons is illustrative of such changes.
(vi) Terminological consistency
Terminological consistency is required when
localization is applied to more than one product. When terminological variations
are necessary, localization software is functionally capable of performing multiple
Phone numbers are different all around the world. For
web localization, it may be necessary to indicate the country code of the phone
Jose, California, the United States, should include the country code +1, the area
code 408, and the number.
(xii) Units of measure

Basically, two major measure systems are in use: the imperial
Functions in computer-aided translation systems
Translation is the translation of ideology, involving
the transfer of thoughts and ideas from one language (the source language) to
property. Copyright and personal data protection regulations, for example, differ
from country to country.
When a software product is to be localized, a localization project will be created
either by an in-house localization group or an outsourced team, which usually
involves the software developer with its in-house translation staff, the software
Functions in computer-aided translation systems
translation tools, such as translation memory and terminology management tools,
plays an essential role. This is due to the fact that both recurrence and repeatability
are common in localization, which involves the translation of user interfaces,
cations, and others.
The project manager works closely with their
client counterparts throughout the project to ensure effective communication.
This is a type of quality control procedure to make sure that issues are quickly
identiÞ ed and promptly resolved.
(iv) Quality delivery
At this stage, timely and quality delivery must be observed.
In-country reviewers may also be asked to assess the quality of the Þ nal product.
After the delivery of the product, follow-up work and maintenance service have
to be provided. Adjustment may also be required, based on feedback on the
Localization procedure
There is hardly a standard procedure for localization (Ca–estro 2005; DiFranco
and Irmler 2003). It depends on a number of variables, such as the purpose of
memory tools, are of particular importance in the localization procedure, as it

Functions in computer-aided translation systems
The Þ rst concerns the change in the languages for localization. Up to the
present, localization has been made predominantly from English into other
major languages, such as French, German, and Spanish. The trend of Ôreverse
localizationÕ, or localization into English, though starting to emerge, is not
Name of the SystemName of the Editing
AcrossCrossDeskNo system Þ
le extension
DŽjˆ Vu X2Editing Interface.
Fluency Translation
Fluency Workspace
(2) same as source format
Translation Editor
SDL-Trados Studio 2011
(project name)
Translation Workspace
(1) Microsoft Word Plug-in
Wordfast Classic
Workfast WorkspaceWord format extension
XTM Workbench
le extension
Yaxin CAT 4.0
Yaxin CAT Platform
Word format extension
Functions in computer-aided translation systems
Platform-dependent environments are plug-in environments that, in turn, can
be classiÞ ed into two different types: text plug-in environment, such as Microsoft
Word; and website plug-in environment, such as Google Translate.
Text plug-in environments
In order to work on a computer-aided translation system with a platform-
dependent environment, a toolbar of the software must be installed in the ÔAdd-
insÕ ribbon of the text processor (e.g., Wordfast Classic for Microsoft OfÞ
The buttons are usually assigned for commonly used functions for translation,
Website plug-in environment
Some software, such as LogoVisa, PROMT, Memsource Cloud, and Translation
Workspace, offer two interfaces.
This is a useful and practical categorization. Even the editor of the web
browser-based computer-aided translation systems can fall into the two catego-
ries. For example, the translation editors of Translation Workspace and Web-
WordSystem are installed in Microsoft Word as text editor add-ins. The
Functions in computer-aided translation systems
editor. The user interfaces of standalone computer-aided translation systems have
parallel translation editors. In most cases, the source text and translation text
are displayed side-by-side or in a two-column table. The left column contains the
segments of the source document, while the right one contains empty places where
users enter translations. Users may scroll up and down to browse the source and

Figure 3.2
MadCap Lingo V9
Lingotek Collaborative Translation Platform, LogoVista, Madcap Lingo, Mem-
source Web Editor, MLTS An-Nakel Al Arabi, OmegaT, OmegaT+, Open Lan-
guage Tools, OpenTM2, PC-Transer, ProMemoria, PROMT, SDL-Trados Studio,
SDLX, SwordÞ sh, Systran, TM-database, Transit, Transoo Editor, Transwhiz,
TraTools, Wordbee, Wordfast Anywhere, and Wordfast Pro.
Features of the interactive editing environments
Fortis Revolution Editor, for example, allows the translator to load multiple
les in a single window so that several Þ les can be simultaneously handled. This
makes it easier to do searching in global tasks.
Functions in computer-aided translation systems
Correctness in numbers and units

Correctness in Spelling
: Atlas, for example, has the function of Assistance to
nd typos in the original text (spell check) (
Post-editing for machine translation systems
Post-editing is deÞ ned as Ôan activity undertaken for the purpose of rendering
(3) The translation is generally faithful to the original.
(4) The translation is generally faithful to the original text, but there are lin-
guistic errors.
(5) The structure and contents of the original work are not well preserved in
(6) The structure and contents of the original are poorly translated.
(7) The translation reflects neither the structure nor the contents of the
original text.
Grammaticality, on the other hand, is about conforming to the standard
Functions in computer-aided translation systems
Stylistic errors involve using language that is inappropriate to the content
and intention of the text. Machine translation outputs are often stylistically
Typological errors refer to the errors in translating the typological conventions
of speciÞ c text types. They serve as a test of the quality of a machine translation
system when using it to translate samples of different text types, such as jour-
nalistic, technical, and literary writings, to see if the outputs are intelligible and
Stages of editing
Editing in the context of machine translation refers to pre-editing a source
text before it is mechanically processed, revising the generated text during
of the translation varies with the end-user. When the
translation is for reference, internal consumption, or concerns basic infor-
for a medium term, such as roadwork notices; and for a long term,
Purposes of a Translation
are varied. Translations may be used for
information purposes. For instance, machine translation has been used
as a tool to collect information on certain areas or regions for political
or intelligence purposes. They may also be used for internal or external
distribution, which require minimal post-editing. However, if the pur-
to be high, the time of production is relatively long, and the revision
work tends to be enormous.

refers to the amount of editing that is required to close up the
Functions in computer-aided translation systems
of delivery. Rapid post-editing, as deÞ ned by Alan Melby, is a Ôstrictly minimal
editing on texts in order to remove blatant and signiÞ cant errors . . . stylistic
issues should not be consideredÕ.
Post-editing issues
There are several lexical issues relating to post-editing.
Mistranslated words
Mistranslated words are words which are translated wrongly by a machine
the approach that the system adopts. Inference can be applied to the output of
a dictionary-based machine translation system.
Words can be wholly mistranslated or partially mistranslated. In the case of
Chinese-English translation, a partially wrong translation may end up being

Unfound words
Local terms, which may not be stored in a system produced in a different country,
are considered unfound words and are usually poorly translated.
Proper-noun translation is extremely difÞ cult for machine translation systems.
Proper names include personal names, object names, place names, group names,
art names, trademark names, and names of historical events. In Chinese, for
instance, a proper noun
may refer to the name of a person, place, or
The regional variations in proper noun translation are problematic. Translit-
eration systems for mainland China (Hanyu pinyin) and for Taiwan (Wade-Giles)
are inconsistent. Similarly, transliteration for proper names in Hong Kong and
In Chinese, there are several ways to translate proper nouns: (1) follow-
ing the original pronunciation of the proper noun (
ming sui zhu ren
), such as Tokyo as
; (2) using the
most popular translation of the proper noun (

such as Churchill as
as Sun Yat-
sen; or (3) following the standard transliteration of the proper noun. In
English-Chinese translation, this means transliterating the proper noun
according to its standard pronunciations, such as putting Belgrade into
Functions in computer-aided translation systems
is called a loan word or phrase. It is fairly common to pronounce the loan word
language readership.
Translation theorists suggest that the connotations of an English summer day
may not be understood by other people as beautiful and hot. Where the source
Functions in computer-aided translation systems
language culture is important, it has to be preserved; where it is not, there is

clariÞ es what the source term means in the translation.

Functions in computer-aided translation systems
The third way (of translation) is that of imitation, where the translator (if
now he has not lost that name) assumes the liberty, not only to vary from
the words and sense, but to forsake them both as he sees occasion; and
taking only some general hints from the original, to run division on the
groundwork, as he pleases. . . . I take imitation of an author . . . to be an
Functions in computer-aided translation systems

explains a source expression in the translation.

language. Literal translation ranges from morpheme-for-morpheme, word-
for-word, phrase-for-phrase, and clause-for-clause translation. The smaller the
unit is, the greater the literalness. Literal translation adheres to the form and
syntactic structure of the original. Thus, this strategy is frequently used in
Functions in computer-aided translation systems
There are two types of omission: general and grammatical. General omissions
To achieve this, it is, as a general principle, always safe to render the expression
literally if it is possible to recognize in context the same intended meaning or
force of the message in the source language. An example of this is the transla-
yi ru fan zhang

into English as Ôas easy as turning over oneÕs
Functions in computer-aided translation systems

involves using a substitute word to replace a word or phrase that
Functions in computer-aided translation systems
words (Feng S. 1981: 7Ð10). The source language word-order is preserved
and the words translated singly by their most common meanings, out of
IsnÕt it also great when friends visit from distant places?
And to have a friend visit from somewhere far away Ð isnÕt that still a
great joy?
The translator of this sentence, after comparing the different translations, can
Functions in computer-aided translation systems
dictionaries, to check spellings. XTM, for example, covers 90 languages. Some
languages, such as Chinese and Japanese, are not spelt, and cannot have spell

makes sure that the same or similar style has been reproduced in
Functions in computer-aided translation systems
Barlow, Michael (1996) ÔAnalysing Parallel Texts with ParaConcÕ,
Proceedings of the
Joint International Conference of the Association for Literary and Linguistic
University of Bergen, Bergen, Norway.
Barlow, Michael (2004) ÔParallel Concordancing and TranslationÕ,
Translating and
, London: The Association for Information Management.
Functions in computer-aided translation systems
avar, Damir, Uwe KŸssner, and Dan Tidhar (2000) ÔFrom Off-line Evaluation to
On-line SelectionÕ, Wolfgang Wahlster (ed.)
Verbmobil: Foundations of Speech-to-
speech Translation
, Berlin: Springer Verlag, 597Ð610.
Champollion, Yves (2006) ÔVery Large Translation Memories: Is the Free Model
Translating and the Computer 28
, London: The Association for Informa-
Chan, Sin-wai (1991) ÔProblems in Philosophical Translation: Translating the Major
Chuang, Thomas C., G. N. You, and Jason S. Chang (2002) ÔAdaptive Bilingual
Sentence AlignmentÕ, Stephen D. Richardson (ed.)
Machine Translation: From
Research to Real Users
, Berlin and New York: Springer Verlag, 11Ð30.
Compiling Group of
A Classified and Illustrated Chinese-English Dictionary
Guangzhou Institute of Foreign Languages
A Classified and Illustrated
Chinese-English Dictionary
), Guangzhou: Guangdong PeopleÕs Publishing
Corbeil, Jean-Claude and Ariane Archambault (eds.) (2006)
Merriam-WebsterÕs Visual
, SpringÞ eld, MA: Merriam-Webster, Inc.
Corbolante, Licia and Ulrike Irmler (2001) ÔSoftware Terminology and Localiza-
tionÕ, Sue Ellen Wright and Gerhard Budin (eds.)
Handbook of Terminology
, Amsterdam and Philadelphia: John Benjamins Publishing Company,
Vol. 2, 516Ð535.
Corness, Patrick (2002) ÔMulticoncord: A Computer Tool for Cross-linguistic
ResearchÕ, Bengt Altenberg and Sylvian Granger (eds.)
, Amsterdam and Philadelphia: John Benjamins Publishing Com-
pany, 307Ð326.
Functions in computer-aided translation systems
(ÔAutomatic Analysis
and Generation Technology of LanguagesÕ),
New Essays
on Machine Translation of Natural Languages
), Beijing: Language Press, 79Ð121.
Studies on Machine Translation
), Beijing: China Translation
Functions in computer-aided translation systems
(ÔTechnology on Chinese SpeechÕ)
SWCL-2002: First StudentsÕ Workshop
guistics, Peking University, Beijing, China, 337.
Gutt, Ernst-August (1991)
Translation and Relevance: Cognition and Context
Oxford: Basil Blackwell.
Guzm‡n, Rafael (2005) ÔLearning for Localisation Tools Training: Importance of
E-learning for TranslatorsÕ,
Translating Today
Teaching and Researching Translation
Hawker, Sara (ed.) (2008)
), Tampa, FL, the United States of
Ion, Radu, Alexandru Ceau
u, and Dan TuÞ
Proceedings of the 5th International Conference on Language Resources
), Genoa, Italy.
Functions in computer-aided translation systems
Text Processing and Computational Linguistics
versity, Soeul, Korea, 306Ð317.
King, Philip and David Woolls (1996) ÔCreating and Using a Multilingual Parallel
ConcordancerÕ, Marcel Thelen and Barbara Lewandowska-Tomasczyk (eds.)
Proceedings of the Conference on Translation and Meaning: Part 4
, Maastricht:
Euroterm, 459Ð466.
Klein, Ernest (1971)
Langlois, Lucie (1996) ÔBilingual Concordancers: A New Tool for Bilingual Lexi-
Proceedings of the 2nd Conference of the Association for Machine
Translation in the Americas: Expanding MT Horizons
), Montreal, Que-
Larsen, Inger (2004) ÔTraining TomorrowÕs LocalisersÕ,
Lau, D. C. (tr.) (1979)
Lazzari, Gianni (2000) ÔSpoken Translation: Challenges and OpportunitiesÕ,
ings of the 6th International Conference on Spoken Language Processing
Lee, Kai-Fu, Hsiao-Wuen Hon, and Raj D. Reddy (1990) ÔAn Overview of the
IEEE Transactions on Acoustics, Speech, and
Lee, Tan, Wai Kit Lo, and Pak Chung Ching (2002) ÔTowards Highly Usable and
Robust Spoken Language Technologies for ChineseÕ, Yu Shiwen (ed.)
of the 2nd China-Japan Natural Language Processing Joint Research Promotion
University, 259Ð265.
Lee, Young-Suk and Salim Roukos (2004) ÔIBM Spoken Language Translation
Proceedings of the International Workshop on Spoken Language
Functions in computer-aided translation systems
Miller, David, Sean Boisen, Richard Schwartz, Rebecca Stone, and Ralph Weischedel
(2000) ÔNamed Entity Extraction from Noisy Input: Speech and OCRÕ,
ings of the 6th Applied Natural Language Processing Conference
, Seattle, Washing-
Miremadi, Seyed-Ali (1992) ÔLanguage UsersÕ Linguistic Creativity: Word-formation
for TranslationÕ, Cay Dollerup and Anne Loddegaard (eds.)
Teaching Translation
Functions in computer-aided translation systems
Pearson, Jennifer (2000) ÔTeaching Terminology Using Electronic ResourcesÕ, Simon
Philip Botley, Anthony Mark McEnery, and Andrew Wilson (eds.)
Corpora in Teaching and Research
, Amsterdam: Rodopi, 92Ð105.
Piao, Scott S. L., Guangfan Sun, Paul Rayson, and Qi Yuan (2006) ÔAutomatic
Extraction of Chinese Multiword Expressions with a Statistical ToolÕ,
of the 11th Conference of the European Chapter of the Association for Computational
), Trento, Italy.
Dictionary for the Analysis of Literary Translation
Department of Comparative Literature, the University of Alberta.
Pux, Simone (2003) ÔImplementing a Client/Server Based Terminology Manage-
Proceedings of the Localization World Conference 2003
Washington, DC, the United States of America.
The Moving Text: Localization, Translation, and Distribution
Amsterdam and Philadelphia: John Benjamins Publishing Company.
Functions in computer-aided translation systems
Schogt, Henry G. (1988)
Linguistics, Literary Analysis, and Literary Translation
Toronto: University of Toronto Press.
SŽguinot, Candace (1994) ÔTranslating and Advertising: Going GlobalÕ,
and TechnologyÕ), Yu Shiwen (ed.)
Language Processing Joint Research Promotion Conference
Institute of Computational Linguistics, Peking University, 318Ð322.
Sun, Le, Youbing Jin, Lin Du, and Yufang Sun (2000) ÔWord Alignment of English-
Functions in computer-aided translation systems
Machine Translation: From Real Users to Research
, Berlin: Springer Verlag,
(ÔVariations in Style and
Literary TranslationÕ),
New Explorations into Translation Studies
Taipei: Bookman Publishing Co. Ltd., 121Ð133.
(ÔOn Georges MouninÕs Views on
Language and Translation
Xu, Jia, Richard Zens, and Hermann Ney (2005) ÔSentence Segmentation Using
IBM Word Alignment Model 1.Õ
Proceedings of the 10th Workshop of the European
Association for Machine Translation: Practical Applications of Machine Translation
Budapest, Hungary, 280Ð287.
Yamamoto, Kaoru, Yuji Matsumoto, and Mihoko Kitamura (2001) ÔA Comparative
Study on Translation Units for Bilingual Lexicon ExtractionÕ,
Functions in computer-aided translation systems
Zhang, Yujie, Ma Qing, and Hitoshi Isahara (2005) ÔA Multi-aligner for Japanese-
Computer-aided translation can be deÞ ned as the use of computer-based tools
to assist human translators during the translation process. As remarked by Bowker
nition. However,
the term Ôcomputer-aided translationÕ is limited to software that is speciÞ
Computer-aided translation tools can be classiÞ
ed in several different ways.
cations is based on the extent to
which the translation process is automated (Hutchins and Somers 1992: 147;
Lehrberger and Bourbeau 1988: 5Ð7). Within the framework of this classiÞ
Computer-aided translation
According to Bowker, McBride, and Marshman (2008: 27), in relatively recent
years, two tendencies have been developing in parallel. Firstly, an ever-growing
number of translators are making use of computer-aided translation systems,
especially of translation memory (TM) tools. Simultaneously, computer users
are increasingly opting for different types of free and open-source software
(OSS). Therefore, the question of incorporating free or low-cost computer-aided
translation tools into the translation process has come to the fore.
There are many reasons for making computer-aided translation tools freely
study, change, improve, or distribute them. Finally, users should consider the
possible limitations regarding the operating systems on which free computer-aided
translation tools work. Although an interesting number of free computer-aided
translation tools are now cross-platform systems, most of them are still intended
for the dominant Microsoft Windows (MS Windows). There are other free
Computer-aided translation
The majority of these programmes are compatible with popular operating
systems, such as Microsoft Windows and Mac OS, but some of them are cross-
platform systems, which can be used in different hardware and software. Araya,
Computer-aided translation
In his review for the magazine
, Sikes (2009: 16) describes Across
Language Server v.5 as a Ôclient-server system that consolidates translation
memory (crossTank) and terminology management (crossTerm) on a central
database server platformÕ. This version includes other elements with the purpose
of assisting the translation process: crossProject and crossCheck, which are also
present in Across Personal Edition; and two source authoring assistance tools,
namely, crossAuthor and crossAuthor Linguistic. Ršsener explains the importance
by integrating authoring tools with a translation memory system using a single
environment, users may beneÞ t, not only from the use of common authoring
les. On top of that, it currently works only with unformatted text, skipping
formulas and images.
Despite its current disadvantages, Anaphraseus is a promising tool with the
potential of becoming more powerful and functional.

Figure 4.1
Main Window of Anaphraseus
Computer-aided translation
Crowdin was born from the concept of crowdsourcing translation. Anastasious
and Gupta (2011: 637Ð638) describes the phenomenon of crowdsourcing as a
Ôlarge group of people who are available and willing to perform a task that an
outsourcer(s) had asked forÕ. In line with this deÞ nition, crowdsourcing transla-
tion consists of a group of people who are eager to translate content submitted
online, usually for low or no payment, and a platform, such as Crowdin, where
the ÔcrowdÕ can work.
According to its website,
Crowdin has a series of useful characteristics. First
of all, it supports all the popular localization Þ le formats and is able to handle
both documents and software projects. It is optimal for webpage translation
and its design suits the demands of multilingual translation projects. Another
interesting feature is that it provides users with reports and information con-
cerning translation activities, translation quotations, analyses of translation
memory databases, and even biographical information of translators. Moreover,
it allows the members of a translation project to disseminate information, con-
duct online communication, and send messages to each other. Last but not
least, it supports up to 109 languages.
A typical workß ow in Crowdin is described by Morera, Lamine, and Collins
(2012: 197) as follows. To begin with, the project must be conÞ
gured either
as ÔmanagedÕ, whereby members of the crowd have to be accepted by a project
manager, or ÔopenÕ, whereby any person can participate. Once the project has
been conÞ gured, the crowd gains access to the source text. At this point, the
Computer-aided translation
solve technical problems but, as Garc’a and Stevenson highlight, many of the
suggestions and wishes of the users were taken into consideration and incor-
porated into the young computer-aided translation system. On top of that,
after purchasing the initial license, users could beneÞ
t from these upgrades
without any additional costs (Garc’a and Stevenson 2010: 16). Based on the
above, it is not surprising that DŽjˆ Vu became one of the most popular
computer-aided translation tools during the 1990s, especially among freelance
translators. Emilio Benito worked as the president and founder of Atril until
ed at allÕ and 5 being ÔExcellentÕ). It is
therefore concluded that, from a general perspective, DŽjˆ Vu seems to be more
satisfactory compared to Wordfast (3.9), SDL-Trados 2006 (3.4), and Trados
EsperantiloTM is a free and open-source translation memory tool that works
as a standalone system. According to its developer, Artur Trzewik, EsperantiloTM
is part of the more extensive programme Esperantilo, which focuses exclusively This broader programme
language, a simple text editor, multilingual user dictionaries, compatibility with
Esperanto into English, Polish, German, Russian, and Swedish; and English and
EsperantiloÕs latest updated version, Esperantilo 0.993, was released in Decem-
ber 2011 and runs on MS Windows and Linux.
Computer-aided translation
German, Greek, Hungarian, Icelandic, Italian, Latvian, Lithuanian, Maltese, Nor-
wegian, Polish, Portuguese, Slovak, Slovenian, Spanish, Swedish, and Turkish.
According to ESTeamÕs website,

Figure 4.2
User Data of ESTeam
Fluency Translation Suite 2011 and Fluency Enterprise Solutions were the
rst server-based editions. The latest server-based version of the software is
Fluency Translation Suite 2013, which provides access to online machine
translation systems, including Google Translate, Microsoft Bing, and APIs for
other machine translation engines. It is remarkable that it implements a ter-
minology management module with large bilingual dictionaries in 35 different
languages. In addition, it has a WYSIWYG interface, displaying the created
document in a very similar way to the end result. Finally, it supports a wide
Computer-aided translation
assurance checking. In 2013, GE-CCT provided English-Chinese and Chinese-
Chinese computer-aided translation.

GlobalSight TMS is an open-source server-based Translation Management Sys-
GlobalSightÕs predecessor, Ambassador TMS, was originally developed by Glo-
balSight Corporation, a company based in Maryland (USA). The company was
later acquired by Transware and, Þ nally, by Welocalize. According to WelocalizeÕs
cial website (, Derek Coffey, now Senior Vice
President of Technology and Development at Welocalize, had a key role in the
process of open-sourcing Ambassador as GlobalSight TMS in 2009, and has
led the development of the product ever since.
GlobalSight TMS offers multi-functional features, such as alignment, machine
translation, project management, terminology management, and user administra-
tion (
machine translation in a low-cost manual way. The explanation they offer about
the reasons for their choice can provide an insight into the beneÞ ts of using
this particular computer-aided translation tool. As quoted from their research
article (Moran and Lewis 2011: 13):
(i) It is free and open-source.
(ii) It is an active project which is undergoing continual improvements.
(iii) It has a sophisticated web services application programming interface (API).
(iv) It is
they emphasize that GlobalSight SaaS is extremely easy to use. Basically, the
client needs to inform the provider of the particularities of the translation process
as well as of the number of users that will be working on the project, and
Computer-aided translation
(1MB). On top of that, some important computer-aided translation tool func-
tions, such as quality checkers and project management, are not present in this
In conclusion, Google Translator Toolkit is more a document-based rather
than a project-based tool. Most experts agree that the aim of this software is
not to contribute directly to the translation industry, but to acquire free-of-cost
bilingual information from its users in order to optimize its machine translation
system (Garc’a 2011: 219; Garc’a and Stevenson 2009: 18). It is, therefore,
more recommended for volunteers and amateurs rather than professional
GTranslator was created by Fatih Demir in 2000 and later developed by Juan
JosŽ S‡nchez Penas, Pablo Sanxiao, and collegues. This computer-aided transla-
Microsystem as SunTrans. Some years later, Waldhšr built on this experience
to create Heartsome, which in 2004 split into two independent companies, one
in Germany and another in Singapore. They continued developing the software
separately, and Araya was then created by the European branch from the earlier
versions of HeartsomeÕs computer-aided translation tool.
In addition to most common functionalities present in other professional
systems (e.g., TMX Editor, XLIFF Converter, Monolingual Terminology Extrac-
tor, Bilingual Terminology Extractor, Server, and Web Server), Araya has two
features which are of particular importance: (i) it supports all languages, includ-
ing double-byte languages and bi-directional languages, without any restrictions
in combination or in the direction of translation; and (ii) it is compatible with
OASIS XLIFF (i.e., XML Localization Interchange File Format), supporting a
full process of localization. In fact, the editing interface of Araya is called Araya
XLIFF Editor.
As aforementioned, Araya includes most translation-memory professional
functionalities, including a TMX editor, a highly sophisticated tool for the
construction and maintenance of translation memory databases and Þ
les based
on the TMX standard. Furthermore, it can be integrated through XML-RPC
support without heavy efforts into existing processes. In the information bro-
chure provided by the company,
XLIFF, and TMX, ArayaÕs server is optimally prepared to integrate and co-operate
Last but not least, Araya is a cross-platform programme, being able to
run in most existing operating systems, including Linux, Solaris, Unix, and
MS Windows.
Computer-aided translation

Figure 4.3
(v) It is possible to synchronize data with another computer so that they
can be used as a pseudo multi-screen. Thus, some information, such as
translation-memory results and glossaries, can be visualized in a second
Computer-aided translation

According to the ofÞ
cial website,
Lingotek Collaborative Translation Platform
has several remarkable characteristics: (i) the translation project management mode
is updated, making the design of the translation workß ow more ß exible; (ii) the
system supports online chats, enhancing communication among the members of a
translation team; (iii) it can be integrated with API systems, such as Sharepoint,
Drupal, Salesforce, Jive Social CRM, and Oracle UCM; (iv) it is integrated with
two machine translation systems, namely Microsoft Bing and Google Translate, and
allows users to plug-in other engines if necessary; (v) it supports .
le format,
translation memory database Þ le formats, such as .
, and terminology Þ
le formats,
; and (vi) it supports 245 languages and sublanguages.
LogiTerm is a product of Terminotix Inc., a Canadian company based in Quebec
and specialized in computer-aided translation. The latest version, LogiTerm 5.2,
released in September 2012, is available in three different versions: LogiTerm
Pro, LogiTerm Web, and LogiTerm Web Extension Module.
Industry StandardsLingotek Community
Professional translator
2500 words/day125,000 words/day
Average cost per word$0.21/word$0.0008/word
Adapted from Vandenberg 2009: 21
Translation Management Platform Edition.

The main features of the Translation Management Platform Edition include
project management, real-time project progression monitoring, simultaneous
communication, and support of about 60 languages. Moreover, the system can
Computer-aided translation
Out of the many features memoQ has (, two of them are
of particular importance:
The latest version of memoQ Server, memoQ Server 2014, was released in
June 2014 and, according to Kilgray, it includes the following new features:

(i) Project templates and workß
ow automation: In this new version, the
Computer-aided translation
Greek, Hungarian, Italian, Polish, Romanian, Russian Spanish, and Swedish. The
le formats that the system can handle, on the other hand, are very limited, sup-
porting only Microsoft Word and Excel, RTF, HTML, XHTML, and XML Þ
Memsource was developed by Memsource Technologies, a company founded
was ofÞ cially released in Germany in April 2002, whereas the latest version, 3.17,
was released a decade later, in 2012.

Computer-aided translation
in turn can only work on MS Windows. Therefore, MT2007 cannot be used in
any operating system other than MS Windows. As for now, each project can have
only one source Þ le, which must have one of the following supported formats:
MS Word 2007 (i.e., .
), MS PowerPoint
), OpenDocument Text (i.e., .
), OpenDocument Spread
management as its core element. The scope of MultiCorpora lies within the idea
of improving the efÞ
ciency of translators. Gerry Gervais, founder of the company,
was serving as manager of a Canadian government translation department when,
in 2006, he proposed the idea of a corpus-based TextBase tool with an innova-
tive leveraging functionality. The new concept was baptized ÔAdvanced Leveraging
Computer-aided translation
OmegaT has the support of a great number of translators. In 2006, a survey
conducted by the Imperial College LondonÕs scholar Elina Lagoudaki revealed
that OmegaT is the most popular of the open-source computer-aided translation
tools. The results of the survey are divided into several groups. The Þ
rst group
regards those professionals that Lagoudaki calls non-TM-users, i.e., those who
do not use a translation memory system but will be willing to try or buy one.
To the question of which computer-aided translation tools had they heard about,
translation memory tool and, more remarkably, being the only open-source
computer-aided translation tool in the list.

Table 4.2
Most Popular TM tools with Non-TM-users
DŽjˆ Vu
SDL-Trados 2006
STAR Transit
Adapted from Table 3 in Lagoudaki 2006: 15

Table 4.3
Top 10 Most Widely Used TM Tools
SDL-Trados 2006
DŽjˆ Vu
STAR Transit
Adapted from Table 7 in Lagoudaki 2006: 20
Computer-aided translation
A second group of respondents comprises the translation-memory users. When
participants in this group were asked about which translation memory tool they
list (including paid computer-aided translation tools).

Finally, among the users of operating systems other than MS Windows,
OmegaT was found to be the second most widely used translation memory
tool, surpassed only by Wordfast.

In the above results, percentage totals may be more that 100 per cent because
respondents were allowed to select more than one tool.
More recently, an unofÞ cial survey in 2010
revealed that, amongst 458
professional translators, OmegaT was used around a third as much as popular
paid systems such as Wordfast, DŽja Vu, and MemoQ.

OmegaT+, Þ rst launched in 2005, is a free server-based computer-aided transla-
in Java, which makes it a cross-platform computer-aided translation tool. The
newest version, OmegaT+ 1.0.M3.1., was released on 23 October 2012. It
DŽjˆ Vu
SDL-Trados 2006
properties (.
), SRT (SubRip), Typo3 (.
XLIFF, XML, XTag, and Windows Resources (
Although it is not a very powerful tool per se, OmegaT+ is most useful when
Computer-aided translation
From the user interface, we can notice that the following functions are avail-
able: Translation Editor, Matches, Machine Translation, Message, Document,
Further signiÞ
cant drawbacks of OpenTM2 are the fact that it is restricted
to the MS Windows operating system and, as of July 2010, the only supported
project Þ le formats were: HTML (UTF-8 and ASCII), JAVA properties Þ
ce Þ
les, XHTML, XLIFF, and generic XML. Fortunately, it supports
the import and export of TMX-format translation memory databases.
enables tasks to be automated, thus allowing translators to quickly create auto-
is an acronym for ÔOpen Source Translation Management SystemÕ,
which is a cross-platform web-based system developed by Klemens Waldhšr and
Rainer Kemmler in Germany. OpenTMS is implemented in Java and requires
Java 1.5 or later editions to operate. It has some basic translation memory
functions, such as pre-translation, concordance search, and conÞ
guration of
matching rates. The processes of OpenTMS involve (i) converting the source
le; (ii) translating the Þ le in the ÔTranslation EditorÕ while consulting terminol-
ogy and translation memory databases; and (iii) back-converting the Þ le to its
original format.
OpenTMS has been discontinued and is now part of the project OpenTM2.
Computer-aided translation
c dictionaries, which could be purchased separately.
(iii) User or customized dictionaries, the most interesting of the three types.
They could be created and modiÞ ed by using the Dictionary Editor tool.
Customized dictionaries enable users to adapt the software to their speciÞ
needs, thus gaining control and speed over both the translation and the
post-editing process.
New versions of the software keep offering these applications, allowing trans-
lators to create their own electronic dictionaries, for manual or automatic use.
In May 2014, the company announced the release of the latest version of
the software, PROMT 10, which is offered in desktop series and corporate
The corporate series encompasses server-based systems, which in turn
are available in several editions: PROMT Translation Server 10, PROMT Trans-
lation Server 10 DE, PROMT Cloud, and Industry Products. Industry Products
refers to server solutions designed speciÞ cally for key industries of the Russian
The computer-aided translation functions of PROMT 10 include all the utili-
ties that were present in the previous version, PROMT 9.5, such as automatic
term and translation memory concordance, term extraction, spelling check,
dictionaries, and terminology management, as well as some new features, such
as a new cross-platform web interface, a corporate mobile application, and
individual customization. Regarding Þ le format support, users can import ter-
formats, and translation
memory documents with .
support of 15,000 corporate customers of all sizes in a broad range of Þ
including some prominent companies, such as Norilsk Nickel, TripAdvisor,
Gazprom, Cisco, Lukoil, and PayPal. The contribution by Beregovaya and
Yanishevsky, both members of the staff at PROMT Ltd., describes PROMT
system deployment at PayPal (Beregovaya and Yanishevsky 2010). PayPal is an
exemplary company in terms of localization efforts. Consequently, prior to
deploying machine translation technology in the localization process, the com-
pany required a number of essential factors, the most important of which were:
to Beregovaya and Yanishevsky (2010: 4), with the preceding customization,
PayPal increased productivity by nearly 30 per cent.
SDL is a British localization and translation technology company founded in 1992
by Mark Lancaster. It has now grown into a listed company with numerous ofÞ
in various countries in the world and a huge impact on the translation industry. In
1995, the company acquired its Þ rst translation memory technology, and since
2001, it has been growing exponentially, mainly through acquisitions of other
Computer-aided translation
SDL TMS, on the other hand, is a translation management system that is
cally designed to handle large and complex localization projects with
Similis (Freelance) is a free translation memory system developed by Emmanuel
Computer-aided translation
nitive. According to Planas (2005: 6), when a Þ rst-generation translation memory
is imported, Similis not only preserves the corresponding segments, but also analyses
supports only European languages (e.g., Dutch, English, French, German, Ital-
ian, Spanish, and Portuguese).
Similis can be used either as a standalone version or as a server edition. In
addition to Similis Freelance, which is now made free, the other versions include:
Similis LSP, Similis Server, and Libellex.
Similis LSP allows users to centralize translation memories and glossaries on a
computer and to share the projects with translators who have Similis Freelance
installed on their computers. Similis Server, in turn, enables a group of translators
Computer-aided translation
Snowman (Free) is a free standalone computer-aided translation system developed
As of August 2013, the latest version, V1.33, provides translations from
Chinese into English, Japanese, French, Spanish, Russian, Korean, and German.
It runs on the latest Windows operating system and supports the following
document formats: plain text (.
) and Microsoft Word/Excel/PowerPoint
2003 (
Snowman (Free) is very restricted regarding the supported formats and languages,
Back in 1995, SyNTHEMA started developing a machine translation system
Computer-aided translation
to the ofÞ cial website (, ensures a higher execution
speed and a smaller programme size. Since Win32 API is a Windows application
programming interface, TM-database is constrained to operate either on MS
Windows NT/2000/XP/Vista/7 or on Linux (by using Wine, which enables
MS Windows applications to run on Unix-like operating systems).
This free
computer-aided translation
system supports merely twelve languages,
namely Bulgarian, English, French, German, Hungarian, Italian, Russian, Span-
ish, Turkish, Chinese, Japanese, and Korean. Although twelve may seem a limited
number of supported languages, it is noteworthy that it includes some character-
based languages, which are generally not supported in most free computer-aided
document formats, including Microsoft ofÞ
ce (.
), Plain Text (.
), Qt sources (.
), PHP sources (.
Android resources (.
), Microsoft resources (.
), Java resources (.
les (.
) formats.
TM-database is a tool aimed at professional translators and, as such, it offers
some interesting features, for instance: (i) customizable segmentation; (ii) trans-
lation memory tool, including fuzzy matching and match propagation; (iii)
glossary and dictionary matching; (iv) translation memory and reference material
searching; and (v) spell-checker.
As aforementioned, TM-database has a terminology tool for searching and
managing terms; however, this tool has not been integrated with the translation
editor. On top of that, this system lacks both terminology management and
project management functions.
Tr-AID, a product of the Institute for Language and Speech Processing (ILSP)
in Greece, was launched in 1995 as the Þ rst translation memory tool developed
by a Greek organization. As quoted from the ofÞ cial website (http://www.ilsp.
gr/traid_eng.html), the aim of this project was Ôto give the Greek and foreign
Translation Workspace is a spin-off of a former computer-aided translation
system called Logoport, which consisted of a real-time translation memory
platform. Logoport started to operate in early 2005, and by 2009 it had handled
more than two billion words and serviced more than 19,000 users and 700
Computer-aided translation
Polish, Portuguese, Romanian, Russian, Serbian, Slovak, Slovenian, Spanish,
Swedish, Turkish, and Ukrainian. Special versions also include Arabic, Chinese,
with a translation memory and a termbase as well as access to a very interesting
resource: the WebWordSystem Public Area. The Public Area consists of a mar-
Computer-aided translation
users who considered themselves to have ÔadequateÕ computer skills, and it was
ranked high in usability.
Wordfast Anywhere enables translation memories
and glossaries to be
uploaded, as shown in Figure 4.14

An online aligner allows users to upload previously translated source Þ
les and
their corresponding translations in order to create a translation memory. In
addition to accessing the translation memories that have been created or uploaded

Figure 4.14
Screenshot of TMs and Glossaries Management in Wordfast Anywhere
WordFisher is a free system developed by a translator named Tibor Kšrnyei
from Hungary, for translators who work in the MS Word environment. The
programme is written in the WordBasic language and it depends entirely on
MS Word, which becomes its main disadvantage. The drawback is not solely
due to MS Word dependence, which has proved not to be a disadvantage for
other popular computer-aided translation tools such as Wordfast Classic. The
problem lies with the fact that WordFisher requires MS Word 6.0 or above to
work, but is not compatible with MS Word 2007 and any other later versions,
WordFisher had some interesting features. For instance, it included an align-
ment option, which many other computer-aided translation tools lack; it allowed
context check within the project; and it was able to automatically build a bilingual
corpus. Corpora could be later uploaded to any translation memory system,
avoiding time-consuming preparatory work. However, Kšrnyei (2000: online)
remarks that, although they can be searched directly by WordFisher, it is recom-
mendable to handle larger corpora by using external programmes.
WordFisher had numerous supporters, but in recent years, the usage of this
computer-aided translation system has decreased compared to other computer-aided

Figure 4.15
Screenshot of the Login Window in Wordfast Anywhere
Computer-aided translation
translation tools. We can conclude that, notwithstanding the advanced features that
WordFisher had when it was Þ rst released, nowadays, it is not as convenient as other
free computer-aided translation tools.

XTM International,
a company headquartered in the United Kingdom, was
founded in 2002 by Andrzej Zydron and Robert Willans to promote their
innovative approach to the translation of XML Þ les: XML:tm (Zydron 2003:
online; Zydron 2004: online). Soon after, in 2004, XML:tm developed into
the Þ rst version of XTM, an XML-based content management and translation
system. In 2005, Lingo24 teamed up with XTM International, investing in
the development of this computer-aided translation tool and integrating it
with Lingo24Õs internal systems. Hazel Mollison, Communications Executive
at Lingo24, published an article listing the advantages that had led to the
implementation of this particular system (Mollison 2007: online). According
to Mollison, the system has a number of key features designed to support
the translatorÕs work, such as an intuitive Java user interface, which displays
and facilitates the processing of different translation memory matches and
supports spellchecking by using customized dictionaries and quality assurance
functions. Mollison also emphasizes XTMÕs extreme scalability for supporting
professional translation services. Based on the foregoing, she concludes that
XTMÕs results are indeed cost-saving at the same time that the overall quality
is increased.
On top of the features mentioned by Mollison, XTM includes a wide range
ow; collaborative projects,
users, and clients; translation memory database and terminology management;

Figure 4.16
Working Document of WordFisher
reports of translation memory analysis and translation fees; integration with
machine translation engines (Google Translate and Asia Online); online editing
mode and quality checking tools; real-time project progress status enquiry; and
plug-in content management systems application programming interface (API).
Furthermore, the system is not limited in terms of Þ
le formats, since it is capable
of handling Microsoft OfÞ
ce, OpenOfÞ
ce, FrameMaker, InDesign (.
), HTML, XML (generic), PDF, RTF, Microsoft Visio (
), Java proper-
ties, DITA, RESX, and TTX formats.
XTM is available in two editions: server-based XTM Suite and XTM Cloud.
XTM Cloud, in turn, has editions for freelancers, small translation teams, and
language service providers. XTM Cloud was Þ
rst released in September 2010
Little wonder then that XTM has secured a growing list of clients, including
numerous Language Service Providers, such as Cuttingedge, Language ScientiÞ

Yaxin CAT
Yaxin CAT is a server-based computer-aided translation system developed by
Beijing Orient Yaxin Software Technology Co., Ltd.

Figure 4.17
XTM Cloud
Computer-aided translation
with the participation of some other companies. In 2005, the development of
Yaxin was performed by Beijing Dongfang Yaxin Software Technology Co. Ltd.
Ye (2010: 180), who provides a thorough evaluation of this system, remarks
that Yaxin CAT combines translation memory with machine translation with
the purpose of aiding translators to achieve higher quality and efÞ ciency in the
translation process while reducing the costs. According to Ye, Yaxin CATÕs
translation memory system consists of three different kinds of memory: glossary
memory and segment memory, both of which are encompassed in the so-called
Machine translation
Machine translation systems are typically divided into two types: commercial
Computer-aided translation
knowledge base that describes the concepts and their relationship inherent in the
application domain. The complexity should be deep enough to allow for all inter-
numerous scholars (Aramaki, Kurohashi, Kashioka, and Kato 2005; Auerswald
2000: 418Ð427; Carl and Way 2003; Somers 1999: 113Ð157).
Interlingua-based machine translation
This is an approach in machine translation whereby translation proceeds in two
different stages: input sentences are Þ
rst analysed into some abstract and ideally
language-independent meaning representation, from which, subsequently, trans-
lations in several different languages can be potentially produced (Cavalli-Sforza,
Czuba, Mitamura, and Nyberg 2000: 169Ð178; Dave, Parikh, and Bhattacharyya
Knowledge-based machine translation
have for use in the computer. Indeed, it is assumed that with the knowledge-
based approach, the computer should be able to disambiguate, process illogical
expressions, and Þ nd implicit meaning from insufÞ cient information (Asaduz-
Memory-based machine translation
Memory-based machine translation refers essentially to a type of machine
translation system that contains the technology of a translation memory
(Hod‡sz 2006).
N-gram-based machine translation
N-gram-based machine translation is an approach in machine translation that
Computer-aided translation
specialized. Further research in this machine translation approach has been car-
ried out by Charoenpornsawat, Sornlertlamvanich, and Charoenporn (2002),
Elming (2006), Proszeky (2005: 207Ð218), as well as Zhu and Wang (2005).
Shake-and-Bake machine translation
The Shake-and-Bake approach is divided into two stages. The Þ
rst stage involves
analysing the source language sentences with the source language grammar, which
produces highly constrained lexical and phrasal signs. The second stage consists of
machine translation engine was a built-in system. Current hybrid systems, on the
other hand, are integrated with online machine translation engines. Eighteen years
after the introduction of hybridity by IBM, this model has become the norm in
translation products. A list of hybrid computer-aided translation tools with their
respective integrated online machine translation systems is provided in Table 4.5

Table 4.5
CAT Systems and their MT Links
Computer-aided Translation SystemsOnline Machine Translation Systems Linked
AcrossGoogle, Moses, Lucy LT, Microsoft
Alchemy PublisherGoogle, PROMT
AutshumatoGoogle, Apertium, Belazar
Computer-aided translation
Machine translation systems used by computer-aided

Table 4.6 provides a list of Þ fteen online translation systems that are linked to
computer-aided translation systems.

funding to build a free and open-source machine translation system for the four
cial languages in Spain. Merely three months later, in July 2004, the Spanish
Ministry of Industry, Tourism, and Commerce agreed to fund the development
of two different machine translation systems: Matxin, for Spanish to Basque;
Name of Online Translation
aided Translation
Google Translate23
Microsoft (Bing)8
Yahoo! Babel Fish
LEC Translate
Lucy LT
SDL Automated Translation1
AppTekÕs TranSphere Machine
supports the translation of 35 languages, including some minority languages,

Figure 4.18
Apertium Translation Interface
Computer-aided translation
AppTek’s TranSphere® Machine Translation
AppTek’s TranSphere® Machine Translation is a hybrid system that combines
Rule-based Machine Translation (RBMT) with Statistical Machine Translation
(SMT). It provides portability across various platforms and operating systems,
being able to run on standalone workstations, servers, and browsers. One unique
feature is that TranSphere can be integrated with AppTek’s automatic speech
recognition product, namely Plain Speech.
Bi-directional Translation

Table 4.8
Language Pairs for which Uni-directional
Translation is Supported in AppTek’s Tran-
Sphere® MT
Uni-directional Translation
T�agalog English
Ur�du English
Gr�eek English
Babel Fish/Yahoo! Babel Fish
Babel Fish was a web translation service based on SystranÕs technology and created,
originally, by AltaVista. The system was named after a Þ ctitious creature in Douglas
. As quoted from the book:
The Babel Þ sh is small, yellow, leech-like, and probably the oddest thing
in the universe. It feeds on brain wave energy, absorbing all unconscious
Computer-aided translation
Original Equipment Manufacturer provider, supplying the core technology to
some of the leading Þ rms in the translation software industry. Some of the
companies that use LECÕs technology in their products are LogoVista, Cross
Language, Panasonic, Babylon, Avanquest, and Softissimo.
LEC Translate is an automatic translation system that can be used to translate
European, Eastern European, and Middle Eastern languages, more speciÞ
Arabic, Brazilian, Chinese, Dutch, English, Farsi, French, German, Hebrew,
Indonesian, Italian, Japanese, Korean, Pashto, Persian, Polish, Portuguese,
Russian, Spanish, Tagalog, Turkish, Ukrainian, and Urdu. However, it must
cantly lower than expected; and, consequently, they had to implement
another machine translation system, Moses. Currently, they use both Lucy LT
and Moses for translations involving Italian. For the other languages, however,
Lucy fulÞ lled the expectations. Lu claims that, by implementing these systems,
CA Technologies dramatically increased the translation output, even to 100 per
Computer-aided translation
cent for some languages, while reducing the post-editing costs considerably (Lu
Lucy Software also offers a free online machine translation tool: Lucy LT
KWIK Translator, which can be accessed at

Microsoft Translator (Bing Translator)
As stated by Wendt, responsible for Bing Translator and Microsoft Translator
services, Microsoft had deployed its own SMT system for internal use since
2002. However, they used to integrate third-party machine translation engines
with their products, and it was not until 2007 that the system became available
to the general public at the Microsoft website (Wendt 2010: unpaged).

Figure 4.19
Lucy LT KWIK Translator
Microsoft Translator offers different solutions adapted to the users:
(i) Collaborative Translation Framework, which started being available in
March 2010.

(ii) Translator Hub, a translation portal and web service that enables to build,
Computer-aided translation
is an open-source project, which incorporates contributions from many sources
and is mainly used at academic institutions as the basic infrastructure for SMT
Computer-aided translation
(iii) the greater the amount of content that is available in multiple languages,
the easier it is for customers to serve themselves through a companyÕs website,
reducing the costs associated to sales and customer service.
SDL also offers instant text machine translation with SDL EasyTranslator,
mainly aimed at students and consumers. SDL EasyTranslator is offered both
as a free version, which can be downloaded from SDLÕs website,
premium version with extended features.
Systran dates to 1968, being among the oldest Þ rms in the machine transla-
Computer-aided translation
Integration of MT with TM Systems
The integration of translation memory into machine translation systems is about
the use of translation units and terminology database in automatic translation.
There are sixteen machine translation systems which have a translation memory
component, namely Atlas, Crossroad, Dr. Eye, EsperantiloTM, Hongyaku,
Computer-aided translation
Al-Onaizan, Yaser and Kishore Papineni (2006) ÔDistortion Models for Statistical
Machine TranslationÕ,
Proceedings of the Joint Conference of the International Com-
), Sydney, Australia.
Alonso, Juan Alberto and Andr‡s Bocs‡k (2005) ÔMachine Translation for Catalan-
Spanish: The Real Case for Productive MTÕ,
Proceedings of the Tenth Conference
on European Association of Machine Translation (EAMT 2005)
Anastasiou, Dimitra and Rajat Gupta (2011) ÔComparison of Crowdsourcing Transla-
tion with Machine TranslationÕ,
Journal of Information Science
37(6): 637Ð659.
Aramaki, Eiji, Sadao Kurohashi, Hideki Kashioka, and Naoto Kato (2005) ÔProba-
bilistic Model for Example-based Machine TranslationÕ, Proceedings of
TranslationÕ, John S. White (ed.)
Envisioning Machine Translation in the Informa-
tion Future
, Berlin: Springer Verlag, 169Ð178.
A Dictionary of Translation Technology
Chinese University Press.
Charoenpornsawat, Paisarn, Virach Sornlertlamvanich, and Thatsanee Charoenporn
(2002) ÔImproving Translation Quality of Rule-based Machine TranslationÕ,
ceedings of the Workshop on Machine Translation in Asia
), Taipei,
Chereshnovska, Marta (2013) ÔTraining for Technical Translators: An Interview with
Uwe MueggeÕ, Available from
Coffey, Derek (2010) ÔWhatÕs Happening with GlobalSight?Õ Online Posting. 19th
March 2010. GlobalSight Community, Available from http://www.globalsight.
Comprendium (2004) ÔComprendium Translator System OverviewÕ, Available from
Condak, Milan (2004) ÔWorkß ow in Wordfast and Invisible Machine TranslatorÕ,
International Journal of Translation
Crego, Josep Maria, Adria de Gispert, and Jose B. Marino (2005) ÔThe TALP Ngram-
based SMT System for IWSLT Õ05Õ,
Proceedings of the International Workshop on
Spoken Language Translation
), Pittsburgh, PA, the United States of
Machine Translation SystemÕ,
International Journal of Translation
Dave, Shachi, Jignashu Parikh, and Pushpak Bhattacharyya (2001) ÔInterlingua-based
EnglishÐHindi Machine Translation and Language DivergenceÕ,
Machine Transla-
Directorate-General for Translation of the European Commission (2005) ÔTransla-
tion Tools and Workß owÕ, Available from
Computer-aided translation
Syntactic Translation ModelsÕ,
Proceedings of the Joint Conference of the International
), Sydney, Australia.
Garc’a, Ignacio (2011) ÔTranslating by Post-editing: Is It the Way Forward?Õ
Garcia, Ignacio and Vivian Stevenson (2009) ÔGoogle Translator Toolkit: Free Web-
based Translation Memory for the MassesÕ,
Garcia, Ignacio and Vivian Stevenson (2010) ÔDŽjˆ Vu XÕ,
21(1): 16Ð19.
Gdaniec, Caludia (1998) ÔLexical Choice and Syntactic Generation in a Translation
System: Transformations in the New LMT English-German SystemÕ, David L.
Farwell, Laurie Gerber, and Eduard Hovy (eds.)
Machine Translation and the
Information Soup
, Berlin: Springer Verlag, 408Ð420.
Gerasimov, Andrei (2001) ÔAn Effective and Inexpensive Translation Memory ToolÕ,
Translation Journal
Gerasimov, Andrei (2002) ÔReview of Wordfast (Windows Version)Õ,
Computing and Technology
Gervais, Dan (2003) ÔMultiTransª System Presentation: Translation Support and
Groenewald, Hendrik and Liza du Plooy (2010) ÔProcessing Parallel Text Corpora
for Three South African Language Pairs in the Autshumato ProjectÕ, Guy de Pauw,
HandrŽ Groenwald, and Gilles-Maurice de Schyrer (eds.)
Workshop on African Language Technology AfLaT 2010
Guerra Martinez, Lorena (2004) ÔPROMT Professional and the Importance of
Building and Updating Machine Translation DictionariesÕ,
International Journal
of Translation
Gu, Liang and Gao Yuqing (2004) ÔOn Feature Selection in Maximum Entropy
Approach to Statistical Concept-based Speech-to-speech TranslationÕ,
of the International Workshop on Spoken Language Translation
IWSLT 2004
Kyoto, Japan.Hearne, Mary and Andy Way (2003) ÔSeeing the Wood for the
Trees: Data-oriented TranslationÕ,
Hearne, Mary and Andy Way (2006) ÔDisambiguation Strategies for Data-oriented
Proceedings of the 11th Workshop of the European Association for
Machine Translation: Machine Translation
Translation Aids

Tools to Increase
, Oslo, Norway, the University of Oslo.
Computer-aided translation
Hutchins, John and Harold Somers (1992)
An Introduction to Machine Translation
London: Academic Press, Available from
Miller, Franois (2002) ÔReview of Wordfast (Macintosh Version)Õ,
Computing and Technology
Mollison, Hazel (2007) ÔLingo24 and XML-INTL Welcome You to XTMÕ, Online
, Available from http://blog.lingo24.
Computer-aided translation
NAACL HLT 2010 Workshop on Computational Linguistics and Writing: Writing
SDL (2013) ÔSDL WorldServer 10.4 Release NotesÕ, Available from http://kb.sdl.
Zhu, Jiang and Wang Haifeng (2005) ÔThe Effect of Adding Rules into the Rule-
To many, computer-aided translation is simply about clicking buttons and saving
data on computer systems for reuse in the future. However, computer-aided

Table 5.1
A Framework for Computer-aided Translation Studies
improvement. It is hoped that this proposed framework will help to organize
concepts and ideas in machine translation and computer-aided translation in a
more coherent and logical manner.
Allen, Jeffery (1999) ÔAdapting the Concept of ÒTranslation MemoryÓ to ÒAuthor-
ing MemoryÓ for a Controlled Language Writing EnvironmentÕ,
Translating and
, London: The Association for Information Management.
In recent decades, translation technology has become increasingly popular both
in Asia and in the West. It is used by professional translators as a core compo-
nent of a personal workstation, by occasional users as an important means of
multilingual information mining, and by international corporations as the foun-
dation of global translation management systems. All the drastic changes resulting
from the use of technology in translation practice in past decades have brought
a revolution in translation. It would not be improper to describe this great
transformation from the traditional to the modern with the words of Zhao Yi
Reading Notes on Twenty-two Histories

), that Ôwhile people are still entrenched
in their traditional thinking, a new milieu has been created by the will of HeavenÕ
The future of translation technology
translation, including both machine translation and computer-aided translation.
We will trace the major happenings in each area to see what lies ahead.

Translation theory: Past and present
The development of translation theory in the world can be divided into four
periods, dominated Þ rstly by the philological approach, secondly, by turns,

Table 6.1
Divisions of Translation
The future of translation technology
The future of translation technology
in Germany, and in Israel, Evan-Zohar proposed the polysystem theory (Even-
Zohar 1978: 21Ð27). In 1980, the deconstructionist theory of translation based
on Jacques Derrida appeared (Derrida 1987), and four years later, Katharina
The future of translation technology
Translators Þ nd that many of the ideas and concepts in the Þ eld are mainly
The future of translation technology
1992: 5Ð19). With the spread of Christianity in the year 30, translation acquired
the new role of disseminating the gospels of Christ.
About 350 years later, in 67, Zhu Falan
Sutra in Forty-two Sections Spoken by the Buddha
tion of a Buddhist sutra made in China. The beginning of large-scale translations
of Buddhist scriptures began in 147 with An Qing
. Since then, the translation
of Buddhist scripture has been an activity involving more than thousands of trans-
lators over a period of 980 years (Editorial Committee 1988: 103).
The case is more or less the same with the translation of the Bible in the
West. It started with the translation of the entire Bible into English by John
Wycliffe in 1382 (Worth 1992: 66Ð70). This is generally regarded as the Þ
The future of translation technology
The future of translation technology
The future of translation technology
In 1955, Japan became the fourth country to develop machine translation.
rst university in Japan to begin research on machine
research on machine translation with a Russian-Chinese translation algorithm
Technology (Dong 1988: 85Ð91; Feng 1999: 335Ð340; Liu 1984: 1Ð14).
Two years later, Charles University in the former Czechoslovakia began to
The future of translation technology
that have emerged in recent decades or even years will continue to exert con-
siderable inß uence on the development of translation technology, which include
the globalization of translation, the creation of multiplicity, the redeÞ
nition of
The future of translation technology
forms of images and various types of sound. As images are beyond words and
beyond cultures, they will be widely used in the future to help tear down the
language and cultural barriers that separate so many language communities in
RedeÞ nition of translation: From the traditional
The Þ rst direction concerns the way translation is to be redeÞ ned in this digital
age. As technology is increasingly used in translation practice and the translation
industry, it is deemed essential that technology be given due emphasis in the
nition of translation.
Translation has been deÞ
ned in a traditional way, depending on oneÕs view
on the goals of translation, the nature of translation, the practice of translation,
and the disciplines related to translation. This has given rise to a number of
nitions which are either general or rooted in concepts before the advent of
technology. A general and typical deÞ
nition of translation in a dictionary, for
instance, would be Ôto transfer from one language to anotherÕ, which means
The future of translation technology
language different from its ownÕ. This approach to translation deals with trans-
lational issues primarily from the perspective of the differences in language
structures. There are scholars who regard translation as a semantic transfer,
holding the view that translation is Ôrendering the meaning of a text into another
approach, some scholars believe that translation is a semiotic transfer. To them,
translation is a process by which the chain of signiÞ ers that constitutes the source-
language text is replaced by a chain of signiÞ
The future of translation technology
translation. The European Union can be cited as an example. In its European
MasterÕs in Translation Programme (EMT), the learning of computer-aided
translation is included in its curriculum. From the side of system developers,
c training in their systems in the form of online tutorials has been provided
by system developers for a long time. Training workshops, particularly online
ÔwebinarsÕ, are organized for different levels of users. User groups have forums
to exchange views on speciÞ c issues on speciÞ c systems. Online forums initiated
by individual translators in the capacity of user groups have become an informal
way of gaining knowledge of computer-aided translation. The online forums
provide a platform for system users in different parts of the world to exchange
their ideas on the strengths and weaknesses of different computer-aided transla-
tion systems and share their experiences on the use of the systems. Translation
agencies often provide training to their in-house or freelance translators so that
The future of translation technology
The process of modernizing the profession of translation will continue into
the future.
Generation of neologisms: From old concepts to new vocabularies
Technology has inß uenced the development of translation studies in various
aspects. It has replaced the old concepts with a new vocabulary based on new
developments in the Þ eld and has given rise to corpus-based translation studies,
Technology has brought new concepts and terms to the Þ eld. In the past,
we talked about loose or even ill-deÞ ned terms, such as sense-for-sense transla-
The future of translation technology
on personal insights. They were therefore mostly empirical, deductive, and
c. In recent decades, concepts and principles in translation have been
formed through the use of translation corpora.
This change is signiÞ cant, as translation studies has shifted from personal
subjectivity to data-based objectivity. The observations and generalizations based
on data will be adequately founded and useful to translation professionals and
scholars. The formulation of concepts will be made in this manner for many
Perfection of automation: From machine translation
The future of translation technology
that with the aid of translation technology, translation is moving towards enter-
prisation. It is moving from profession to business.
Emphasis of practicality: From the literary to the practical
eld is
the growing attention given to the translation of practical texts. In the past,
literary translation was dominant. Both in China and in the West, the translation
of religious documents, such as Buddhist scriptures in China and the Bible in
the West, and classics dominated the world of translation. Nowadays, we translate
mostly practical texts and webpages. According to estimation, practical texts
The future of translation technology
Change of approach: From reactive to proactive
The future of translation technology
The future of translation technology
Delisle, Jean and Judith Woodsworth (eds.) (1995)
Translators through History
Amsterdam and Philadelphia: John Benjamins Publishing Company and UNESCO
Derrida, Jacques (1987) ÔDes tours de BabelÕ,
PsychŽ: Inventions de lÕautre
, Paris: GalilŽe.
Dong, Zhendong (1988) ÔMT Research in ChinaÕ, Dan Maxwell, Klaus Schubert,
and Toon Witkam (eds.)
New Directions in Machine Translation
, Dordrecht-
A Dictionary of Translators in China

A Dictionary of Translators in China
Beijing: China Translation and Publishing Corporation.
Even-Zohar, Itamar (1978)
The future of translation technology
Loh, Shiu-chang (1975) ÔMachine-aided Translation from Chinese to EnglishÕ,
United College Journal
Loh, Shiu-chang (1976a) ÔCULT: Chinese University Language TranslatorÕ,
can Journal of Computational Linguistics, MicroÞ
Loh, Shiu-chang (1976b) ÔTranslation of Three Chinese ScientiÞ
c Texts into English
abbreviation 79, 81, 89, 96, 114, 131;
chatroom 130
accessibility 35, 106, 168
acronym 4, 30, 79, 131, 199, 266;
semantic 87, 107, 249; sentence 9;
sentential 87; source 87, 249;
documentary 262, 271; text 30, 87,
memory 215; usage 209
An-Nakel Al-Arabic 9, 11; MLTS
Apertium 172, 194, 221–3
system 253; computer-aided
translation software 17, 167, 203;
computer-aided translation studies
244; computer-aided translation
suite 210; computer-aided translation
201, 209, 213, 222, 226, 233, 247–52,
254, 270, 272–3; computer-aided
207–14; corporate 103; cross-
platform 46, 196; cross-platform
custom-speci c 103; example-based
103;  le-based 105; free 168–9, 171,
CULT (The Chinese University
Language Translator) 266
culture 90, 110, 115–16, 133–4, 138,
142, 261, 268, 274; receptor 134;
source 133–4, 139; source language
104; task-speciÞ c 97; Technical 95;
thesaurus 89; translation memory 98;
disambiguation 249Ð50; reference 217;
word 220; word-sense 249
225, 227; blurred 74; creased 74;
264; documentation format 17, 20;
document format 7, 11Ð12, 42,
document formatting 174; document
95; pre-translated 59; printed 73,
248; project-based 76; project-
speciÞ c 76; religious 273; software
119; source 24, 76, 124; tagged 106;
terminology 200; TMX 107, 197;
184Ð5, 188; translation memory 200
93; domain expert 78; specialized 230
Dostert, Leon 2
Dr. Eye 123, 170, 234
Dryden, John 136, 259
edit distance 129; data 34, 68, 120, 125,
249; editing 33, 42, 73, 120Ð1, 126,
128, 129, 132, 216, 253; editing
environment 34, 78, 121, 123Ð4;
editing platform 99, 121; editor 5, 88,
106, 123, 209; full 129; interactive 34,
40, 120, 125, 129, 132, 253; manual
technical 4; maximum 129; minimal
129Ð30; multi-format translation 171;
online 215; online translation 232;
parallel translation 123Ð4; platform-
dependent 121; platform independent
121; post 34, 37, 40, 126, 128Ð30,
132, 181, 200, 251, 253; pre 40, 52,
formal 140; functional 134, 142,
94; nil 139; of cial 94; one-to-
head-driven phrase structure 249;
249; phrase structure 249; simpli
52, 249; source language 136–7, 220;
Környei, Tibor 8
Lancaster, Mark 7
automatic 72; bi-directional 183;
character-based 208; controlled 52–5,
Indo-European 36, 51, 111; input
55; intermediary 137; ISO 639 196;
ISO-639–1 179; kindred 137, 272;
background 38; language code 49,
117, 119; localization group 117;
localization industry 109–10, 119,
198; Localization Industry Standards
memory store 9; localization plug-in
10; localization process 117, 189,
200, 202; localization project 109,
project management 117; localization
project team 116; localization
software 9, 113–14; localization
work ow 20, 118, 202; post-
localization 118; pre-localization 117;
reverse 120; software 108–9, 111–13,
Locke, William N. 1
Logic-Based Machine Translation
LogiTerm 42–3, 47, 123, 186;
LogiTerm 5.2 21, 186; LogiTerm
Pro 186; LogiTerm Web 186–7;
LogiTerm We Extension Module 186
Logoport 11, 195, 209
LogoVista 11, 123, 125, 170, 227, 234;
LogoVista PRO 2013 22
LongRay CAT 47, 123, 186, 221;
LongRay CAT 3.0 19, 186
Lucy LT 221–2, 227; Lucy LT KWIK
Translator 228
Luther, Martin 263
103, 175, 177, 218; free 231; fully
based 219; internal 205; knowledge-
233; machine translation industry
126; machine translation provider
machine translation services 231;
memory-based 105, 107, 218–19;
221, 233; open-source 222; Pangloss
Mark III 218; rule-based (RBMT)
230; statistics-based 218; third-party
228; transfer-based 105, 220, 227;
translation memory-based 107; word-
for-word 206
Manson, Andrew 16
grammatical 136; lexical 136; part-
Martin, Jacky 261
206, 220; character-string-based
deep 193; dictionary 208; dual 209;
232, 252; glossary 208; greatest
Match 100; linguistically enhanced
176; low-score 177; machine
translation 122, 181; matching
34, 99, 106, 133, 138, 170, 216,
218, 252; matching percentage
179; matching rate 199; match
propagation 194, 208; match rate
Moumin, Georges 135, 260
Multiconcord 87
Multi-Lingual Translation System
(MLTS) 107, 170, 234
MultiTerm 5, 13; MultiTerm 2 6;
MultiTerm Dictionary 13; MultiTerm
Lite 13; MultiTerm Pro 13;
MultiTerm Professional 1.5 8
MultiTrans 7, 12, 41–4, 47, 103, 123,
192, 195, 221; MultiTrans 3 10,
13; MultiTrans 3.5 11; MultiTrans
3.7 14; MultiTrans 4 14; MultiTrans
Prism 21–2; MultiTrans Prism
5.5 21; MultiTrans Prism 5.6
193; MultiTrans Prism Blue 194;
MultiTrans Prism Enterprise Server
193; MultiTrans Prism Red 194;
MultiTrans Prism Yellow 194
Munday, Jeremy 244, 246
name 48, 70, 80, 82, 84, 113, 127,
130–1, 136–7, 172, 222, 225–7;
art 131; bilingual 250; command
111; file 41; geographical 143;
group 131; login 86; multilingual
250; name entity 250; name
transliteration 139; object 131;
personal 130–1; place 130–1,
140; project 71, 121; proper
127, 130–1, 134, 137, 143, 250;
trademark 131; translated 139; user
PC-Transer 125, 234
cross-platform 183; desktop
publishing 43; external 213; freelance
168; internationalized 111; Java 194,
presentation 42; professional 16;
programme code 110; programmer 7,
16, 111, 168; programme size 208;
programming 5, 251; programming
guideline 111; programming language
201, 269; translation memory 8;
Sakhr CAT Translator 104, 170
Savory, Theodore H. 260
Automated Translation 221–2,
EasyTranslator 232; SDL Edit 11;
SDL Maintain 11; SDL MultiTerm
18, 78, 201; SDL MultiTerm 7
Extract 78; SDL MultiTerm Convert
Passolo Essential 18; SDL Project
Wizard 11; SDL Server 2009 19;
SDL Studio GroupShare 201; SDL
Studio GroupShare 2014 SP1 201;
SDL Termbase 11; SDL TMS 201–2;
2013 201; SDL-Trados 41–4, 47,
SDL-Trados 2006 15, 177, 195–6;
SDL-Trados 2007 Synergy 15–16,
18–19, 59–60; SDL-Trados 2015
58; SDL-Trados Studio 125, 201,
229, 231; SDL-Trados Studio 2009
18, 20; SDL-Trados Studio 2011 24,
80, 121; SDL-Trados Studio 2011
Alignment Editor 82; SDL-Trados
Studio 2013 84; SDL-Trados Studio
2014 68, 71, 201; SDL-Trados
WinAlign 6, 9, 12–13, 18; SDL
Web ow 10; SDL Workbench 10,
125; SDL World Server 24, 180,
201–3, 267; SDL World Server 2011
14; SDLX 2005 15; SDLX AutoTrans
11, 221; SDLX Translation Suite 4 11
search 8, 21, 75, 79, 216, 260;
Snowman 41–3, 47, 88, 121, 170, 206,
221; Snowman 1.0 19; Snowman
1.27 22; Snowman 1.3 71, 80, 84;
Snowman Collaborative Translation
Platform 22; Snowman V.1.33 68, 206
software 2, 7–10, 13, 16, 22, 36,
224; computer 167; computer-aided
translation-related 101; controlled
language 54; cross-platform 46;
foreign language 115; free 168; free
open-source 223; localized 98, 119;
machine translation 199; open-source
(OSS) 168, 179; paid 169; rule-based
199; software application 117, 119;
Software-as-a-Service (SaaS) 20, 180,
185, 207, 211; software developer
110–11, 117, 120, 207; software
documentation 119; software
engineering 109; software license
translation software 5–7, 9, 12–13,
36, 77, 246; translation software
industry 227
222, 230, 233, 266; open-source
106; server-based 69, 169–70, 201,
system commercialization 4–5; system
development 6, 168; system provider
103; system requirement 69; system
unidirectional 217; web-based 253;
web-based customer-speci c 108; web
Systran (System Translation) 14, 22, 47,
Systran 7 Premium Translator 232;
Server 5 233
Taber, Charles 31–2
tag 21; ALT (alternative tag) 35; IMG
 y 10; part-of-speech 253; statistical
253; TagEditor 14; tagger 253;
tagging 249, 253; tag protection 6,
terminology extraction system 12;
terminology maintenance module 8;
terminology management 6, 11–12,
terminology management system 77,
105, 118, 198, 252; terminology
194, 198; terminology processing
76, 252; terminology recognition
76, 172, 252; terminology report
79; terminology system 39,
171; terminology tool 14, 208;
terminology translation 76, 252;
Terminology Wizard 207
TermStar 6; TermStar NXT 21
261; alternative text 35; automatic
248; comparative 248; controlled
52; controlled language 52, 253;
controlled source 52; controlled
Trados 4–10, 12–15, 25, 86, 108,
176–7, 195–6, 199, 203; Trados
5.0 10; Trados 5.5 10; Trados 6 11;
Trados 6.5 11–12; Trados 7 Freelance
14; Trados Workbench 11, 191
Tr-AID 168, 208; Tr-AID 2.0 13, 208
TransAssist 107
TransCheck 146, 253
128, 134, 140–1, 220, 269; direct
134; ideational 269; information
approach 107; transference 93, 143;
transfer process 32
TransFlow 18
Transit 4–5, 41–4, 47, 55, 125, 188,
209, 221; STAR Transit 195; Transit
1.0 5–7; Transit 3.0 13; Transit NXT
18, 209; Transit NXT Service Pack 3
20; Transit NXT Service Pack 4 21;
Transit NXT Service Pack 6 23
TranslateCAD 44
text-to-graph 267; text-to-picture
114; webpage 33, 227, 271; word
143; word-for-word 104, 137, 139,
Translation Brain 47, 234
translation management 102; computer-
TransSearch 42
Trans Suite 2000 7, 12–13, 221
TransType 105
Transwhiz 7, 12, 47, 105, 125, 170,
234; Transwhiz 9.0 15; Transwhiz 10
17; Transwhiz Power 11
TraTool 14, 47, 125
Tsutsumi, Yutaka 5
Tyndale, William 263
Tytler, Alexander Fraser 260
143; lexical 94, 143; multiword
90; operational 142; prede
ned 81;
service 180; source language 140;
WebCat 207
WebTerm 6
WebWordSystem 46–7, 123, 210;
WebWordSystem Public Area 211
Wilss, Wolfram 31–2
word 5, 8, 16, 22, 36, 53, 55, 75–9,
120, 127, 130–2, 134–43, 145,
267–8; ambiguous 56; appreciative
92; approved 53; borrowed 139;
isolated 75; loan 133, 139; measure
new 60; Not Translated (NTW)
95; orthographic 144; restricted
171; selected 79; single 77; source-
language 94, 139, 143–4; source
language cultural 94, 133; source-text

Приложенные файлы

  • pdf 10823759
    Размер файла: 3 MB Загрузок: 0

Добавить комментарий