2Rationale for IT Applications to Translation “A computer is a device that can be used to magnify human productivity. Properly used, it does not dehumanize by imposing its own Orwellian stamp on the products of human spirit ……….………..Translation is a fine and exacting art, but there is much about it that is mechanical and routine, if this were given over to a machine, the productivity of the translator would not only be magnified but this work would become more rewarding, more exciting, more human.”Martin Kay (1987)
3COURSE OVERVIEW ESSENTIALS TEXT PROCESSING MT TM WORKING WITH CORPORA TERMINOLOGY EXTRACTION AND GLOSSARY PRODUCTION (MONOLINGUAL AND BILINGUAL CORPORA)
4COURSE OVERVIEW - DETAILS 1) ESSENTIALS:Types of computer aidesCAT vs. MTHistory of CAT toolsGeneral principles of working with CAT toolsReference materialsLocalization and internationalizationUNIXSOME OF THIS TODAY!
5COURSE OVERVIEW - DETAILS 2) TEXT PROCESSINGWord and WordPad (tips and tricks)Fonts, code pages, keyboard layout, language tools in Windows XP and OfficeSpeech recognition softwareScanningOCRFile types (essential info on the most common file types and file conversion utilities)
6COURSE OVERVIEW - DETAILS 3) MTHow it works, brief exhibition:Systran ProPromptNeuro TranBabelfishDESKTOP BASEDSUPPORTS CROATIAN (partially Serbian)WEB BASED
7COURSE OVERVIEW - DETAILS 4) TM:Overview (what it is, standards and file formats)Desktop vs. server based TM programsWinAlignWordFastTrados (nowadays SDL Trados) – Freelance editionSisulizer
8COURSE OVERVIEW - DETAILS 5) WORKING WITH CORPORAEssentialsConcordancing (WordSmith, Concordancer, AntConc)Advanced corpora analysis: WordSmith, TigerSearchLemmatization and annotationParallel corpora: ParaConc
9COURSE OVERVIEW - DETAILS 6) TERMINOLOGY EXTRACTION AND GLOSSARY PRODUCTIONEssentialsDoing it automatically: Trados (i.e. SDL) MultiTerm (Desktop and Extract)Doing it semi-automatically: ParaConc, Concordancer
10COURSE REQUIREMENTS Basic computer literacy Positive outlook: Computers don’t biteCAT tools are not complex, they are actually made to make you more efficientInterest in translationWillingness to become several times more efficient in doing translations
11SCHEDULE HONESTLY, WE DON’T KNOW FOR CERTAIN! THAT’S WHY WE NEED YOUR ADDRESSES, SO THAT WE CAN KEEP YOU UPDATED WITH THE LATEST SCHEDULE DEVELOPMENTSPROBABLY: LOCATION: 25 (lectures) and 38 (computer lab), SATURDAYS, at16:00 O’CLOCK
12LITERATUREGeoffrey Samuelsson-Brown, A Practical Guide for Translators (Topics in Translation), Multilingual Matters, 4th edition (May 28, 2004)H. L. Somers (Editor), Computers and Translation: A Translator's Guide (Benjamins Translation Library, 35), John Benjamins Publishing Co, 1st edition (May 2003)Bert Esselink, A Practical Guide to Localization (Language International World Directory), John Benjamins Publishing Co, Revised 1st edition (September 2000)Silvia Pavel and Diane Nolet, Handbook of Terminology, Translation Bureau of Canada, 1st edition (2001)Frank Austermuhl, Electronic Tools for Translators (Translation Practices Explained), St. Jerome, 1st edition, (April 2001)
13COURSE OVERVIEW - GRADING This is a hands-on courseYou will be graded on the basis of the results of your practical assignments:Creating TMs from parallel texts (fiction and non-fiction e.g. a book and a manual) – in a way, you will be also creating a parallel corpusTranslating two short passages (fiction and non-fiction) using your newly created TMs
14ESSENTIALS AND MORE ABOUT THE COURSE IT AND TRANSLATIONESSENTIALS AND MORE ABOUT THE COURSE
15TYPES OF COMPUTER AIDES Computer aides / tools that are relevant to translators can be roughly classified into three groups:Basic input and editing toolsReference toolsProductivity toolsWORD PROCESSORSElectronic books (desktop & web) Electronic dictionariesWeb (Eurodicautom, onelook, etc.) Software-based reference materials (encyclopedias, e-Bible, etc.)TM toolsMT toolsSpeech Technology (i.e. voice recognition)
16CAT vs MTAs soon as you start using computer software in the process of translating, you are entering the realm of COMPUTER-AIDED TRANSLATION, or CAT in short.In other words, CAT is a form of translation wherein a human translator translates texts using computer software designed to support and facilitate the translation process.
17CAT vs MT (continued)The problem is that COMPUTER-AIDED TRANSLATION, is sometimes also called COMPUTER-ASSISTED TRANSLATION, MACHINE-AIDED TRANSLATION or MACHINE-ASSISTED TRANSLATION.Due to the latter two terms, CAT is sometimes confused with MACHINE TRANSLATION, or MT in short.
18CAT vs MT (continued)Although these two concepts are related and similar in some aspects, CAT and MT denote two diametrically different processes:In CAT, the computer program merely supports the translator, so the translator translates the text himself/herself, making all the essential decisions involved.In MT, the translator supports the machine, that is to say: the computer (i.e. program) translates the text, which is then edited by the translator, or, in most cases, not edited at all.
19Translation Technology Continuum CAT vs MT (continued)Graphically represented, the difference is:Translation Technology Continuumautomationhuman involvementComputer-aidedTranslation (CAT)AutomaticTranslation/Machine TranslationUnaidedTranslationTranslation processautomated by use ofMachine TranslationTranslation process aided by electronic tools such as (most typically) Translation MemoryTranslation processnot aided by anyelectronic toolsAdapted from Hutchins & Somers (1992)
20CAT – its scopeWRONG!!!CAT is traditionally associated with large-scale / corporate translations:manuals and technical documentationsoftware localization“Typewriter-assisted” (i.e. traditional) translation is usually associated with small-scale / individual translations (done by freelancers):fiction books, scientific papers, etc.
21CAT – its scope (continued) This is notion of CAT being restricted to corporate translation projects dates back to the 90s and is based exclusively on financial criteria:during the early and mid 90s a combination of a high-end computer and a high-end CAT tool cost as much as a new carfrom their very beginnings CAT tools were designed to be capable of handling both big- and small-scale projects, but initially no freelance translator could afford them
22CAT – its scope (continued) Even for a freelance translator, CAT route is nowadays the only possibility if one wants to provide high-quality, 100% terminologically consistent and efficiently produced translations.A testimony to that is the industry-standard TM program Trados: Trados Freelance edition has been the company’s best-selling TM program for a number of years.
23CAT tools – a bit about their history CAT tools were developed after (very) disappointing initial experiments with MT tools.So, in order to give you a proper overview of how we got where we are now, we have to start with the history of MT tools
24MT History – how we switched to CAT MT research began in 1950’s – Warren Weaver’s 1949 Memo:“When I look at an article in Russian, I say: This is really written in English, but it has been coded in some strange symbols. I will now proceed to decode.”(in Locke and Booth 1955:18)
25MT History – how we switched to CAT Initially based on some misconception about human translation:knowledge of two language systems sufficesit is merely a matter of looking up dictionariesit is easy to define “a good translation”there is only one correct translation possible
26MT History – how we switched to CAT MT history milestones: pre-ALPAC1954: Georgetown system demosuccessful translation of 49 Russian sentences into English: $50m spent in 20 research centres in USA1966: Automatic Language Processing Advisory Committee (ALPAC) Report concludes:”...MT is slower, less accurate and twice as expensive as Human Translation...”“...there is no prospect of useful MT either immediately or in the future...”
27MT History – how we switched to CAT MT history milestones: post-ALPAC1969 – privately funded projectsLogos system (1969); Weidner-CAT (1977); ALPS (1980)1975 – Météo project in Canada1976 – European Commission acquires Systran1979 – Eurotra project in Europe for Multilingual system1980 – PC-based system1990 – data-driven system; WebMT
28MT History – how we switched to CAT 1975 Météo project in CanadaAutomatic translation of weather forecasts (En -> Fr)Sublanguage approach (domain-specific MT)Most successful MT application to datepublic broadcasting since 1977Fr -> En available since 1989only 4% of output needs post-editingrapid translation staff turnover no longer a problem
29MT History – how we switched to CAT Renewed interest in MT in late 80s and early 90s:Technological factorsspecifically: prevalence of PC with improved processing powerTranslation market factorsofficial bilingualism/multilingualism create institutional needsglobalisation creates huge commercial needsAdvances in computational linguisticsMore realistic user expectationsInternet creates casual access to multilingual information
30MT History – how we switched to CAT However, translations produced by MT were still not reliable and accurate enough for large-scale commercial applications.So, it became evident that the human translator cannot be eliminated and replaced by computers.Actually, it became obvious that computers programs should be used as TOOLS which only HELP the translator.
31History of CAT ToolsUnreliability of MT tools -> large corporations hire translation agenciesTranslations agencies find it difficult to cope with the increasing demandTranslation agencies develop their own in-house CAT toolsTranslation agencies begin to sell their CAT tools
32History of CAT ToolsTwo major players in the domain of CAT tools development Trados and STAR Group both started as:TRANSLATION AGENCIES!!!STAR AG was founded as a small translation agency in 1984 by Josef Zibung and Hanspeter Siegrist in the northern Swiss city of Stein am Rhein near Schaffhausen. It won and keept customers from the automotive, machine tool, computer and aeronautics industries like ABB, AT&T, BMW, Dornier, IBM, Mazda, Mercedes, Nissan, Saab and Siemens.TRADOS was founded in 1984 by Jochen Hummel and Iko Knyphausen in Stuttgart, Germany to provide translation services for IBM.
33TRADOS – timelinefirst version of TRADOS's main component, MultiTerm was created for DOS TRADOS developed the first MultiTerm for Windows (v3.1) 1992 – TRADOS’s Translator's Workbench with linguistic fuzzy-matching on translation memories for DOS TRADOS’s Translator's Workbench for Windows
34TRADOS – timeline (continued) 1997 – BREAKTHROUGH : Microsoft decides to base its internal localization memory store on TRADOS 1998 – Microsoft acquires a share of 20% in TRADOSTRADOS becomes a de-facto industry standard CAT tool!!!That’s why we will mostly work with TRADOS in this course (as far as TM is concerned).But we will also work with WordFast, because not all people can afford Trados.
35WHAT WE WANT TO TEACH YOU HERE? TWO PRACTICAL EXAMPLES OF COMMON TRANSLATION PROBLEMS
37IMPORTANT THINGS TO NOTE: (quite obvious) the book has an index = YOU (i.e. the translator) are supposed to make it in the translated version of the booka vast index = a lot of terminologysome index terms appear on several pages that are not necessarily in the same chapter (e.g. pg. 36, pg. 92 and pg. 255) = a very serious problem for the consistency of you translation
43General principles of working with CAT tools The main goals are EFFICIENCY and CONSISTENCYCAT tools = TM tools (in this case only)The basic idea is fairly simple:Documents, especially technical ones, contain a large amount of content that is similar or identical to information already contained in earlier versions or similar documents that have been translated before.that applies to the source editing language (SL) as well as the target translation languages (TL).
44General principles of working with CAT tools So, wouldn’t it be great to re-use previously translated content as valuable reference material for new translations as well so as to obtain consistency of terminology and phrasing?That is exactly what CAT tools do!CAT tools make it possible for translators to work only on content that is being created for the first time. Existing text and text similar to existing text is taken from the available. reference translations (i.e. from TM= translation memory).
45General principles of working with CAT tools So, wouldn’t it be great to re-use previously translated content as valuable reference material for new translations as well so as to obtain consistency of terminology and phrasing?That is exactly what CAT tools do!CAT tools make it possible for translators to work only on content that is being created for the first time. Existing text and text similar to existing text is taken from the available. reference translations (i.e. from TM= translation memory).
47A DREAM COME TRUE? NOT REALLY TO ENJOY ALL THE BENEFITS OF CAT TOOLS FIRST YOU HAVE TO CREATE A TM AND A TERMINOLOGY DATABASE:either from your old translationsor from new translations (i.e. creating a TM from scratch)A DREAM COME TRUE?THAT IS WHERE OTHER CAT TOOLS (i.e. NON-TM CAT tools) STEP IN TO SAVE THE DAY!!!NOT REALLY
48REUSING YOU OLD TRANSLATIONS The best way to make a TM:reliable source (YOU did the translation)readily available (stored on you PC)
49A BRIEF DIGRESSIONThe term LOCALIZATION has often popped up in previous slidesWhat is LOCALIZATION?
50WHAT IS LOCALIZATION?Localization is the process of adapting, translating and customizing a product (software) for a specific market (for a specific locale or cultural conventions; the locale usually determines conventions such as sort order, keyboard layout, date, time, number and currency formats). In terms of software localization, this means the production of interfaces that are meaningful and comprehensible to local users.The Localization Industry Standards Association (LISA) defines localization as: “Localization involves taking a product and making it linguistically and culturally appropriate to the target locale (country/region and language) where it will be used and sold.”Typically, this involves the translation of the user interface (the messages a program presents to users) to enable them to create documents and data, modify them, print them, send them by , etc.)
51LOCALIZATION – what it includes Focal points of internationalization and localization efforts include:Language:Computer-encoded textAlphabets/scripts; different systems of numerals; left-to-right script vs. right-to-left scripts. Most recent systems use the Unicode to solve many of these character encoding problems.Graphical representations of text (printed materials, online images containing text)Spoken (Audio)Sub-titles for videoDate/time format, including use of different calendarsFormatting of numbers (decimal points, positioning of separators, character used as separator)Time zones (UTC in internationalized environments)CurrencyImages and colors: issues of comprehensibility and cultural appropriatenessNames and titlesGovernment assigned numbers (such as the Social Security number in the US, National Insurance number in the UK) and passportsTelephone numbers, addresses and international postal codesWeights and measuresPaper sizesDifferences between local standards (e.g. YU ISO or JUS) and international standards (ISO)
52LOCALIZATION vs. INTERNATIONALIZATION The distinction between internationalization and localization is subtle but important:Internationalization is the adaptation of products for potential use virtually everywhere, whilelocalization is the addition of special features for use in a specific locale.The processes are complementary, and must be combined to lead to the objective of a system that works globally.
53CAT tools for localization Over the last couple of years, in addition to general-purpose TM tools such as Trados and Transit, translation technology companies also developed a number of TM tools specially designed for localization:Alchemy CATALYSTPASSOLOSisulizerSISULIZER is currently the industry standard localization tool, so that’s the one in which we will work!!!
55Other CAT tools (non-TM based) As we said earlier, computer-assisted translation (CAT) is a broad and somewhat imprecise term covering a range of tools, from the fairly simple to the more complicated, which can include:Word processors, grammar and spell checkers, terminology managers, eBooks, eDictionaries, full-text search tools, concordancers, web, TM tools, bitexts, etc.
56CAT - REFERENCE MATERIALS Reference materials are the primary source of terminology in absence of translation memory.Computer-based reference materials can be classified into:Online librariesSpecialized web resourcesSpecialized software productsOther materials in electronic formats
57Online Libraries Large collections of books in electronic form, e.g. eBrary (new scientific books, pay site)Internet Archive (hosting “A Million Book Project”)Project Gutenberg (PD fiction books, free)Questia (popular titles – fiction and non-fiction, pay site – some sections free)
73Specialized software products Various programs that can be used for terminology extraction:Electronic dictionariesGeneral monolingual: e.g. OED v3Specialized monolingual: e.g. Cambridge Pronouncing Dictionary, Collins CollocationsBilingual: e.g. Morton Benson, MidiDictElectronic Bible (e.g. e-Sword)Concordance programs (e.g. Concordancer)Data-mining programs (e.g. Summarizer Pro)
76Concordancers Make it possible to see a word in context: Two types: Useful for finding collocations and phrasesUseful for extracting terminologyTwo types:Monolingual concordancers (e.g. WordSmith)Polylingual concordancers (e.g. ParaConc)