1	Computer Applications for Translators: Basic CAT Tool Components Sue Ellen Wright Kent State University Institute of Applied Linguistics © Sue Ellen Wright & Alan Melby 2005
2	Computer Applications for Translators Basic text production tools Controlled authoring tools Corpus management tools Terminology management tools Translation “workbench” type tools Translation memory Tagged text protection Machine translation systems
3	Basic text production tools Word-processing software Word HTML-editors Dreamweaver, Homesite Text editors Notepad, Wordpad, Ultra-Edit Desktop publishing systems Quark Express, Framemaker, Pagemaker
4	Processing Order Text/content production (customer side) Corpus compilation Text alignment of legacy files Term extraction Terminology management Translation using integrated translation tools TM + WP + TermBase Project & workflow management (ongoing)
5	Organization of the Eight Tool Functions
6	Corpus Management Tools Concordancers Statistical analyses Global search capabilities Frequency reports XML-based intelligent searches
7	Sources of Corpus Components Enterprise information resources Existing parallel texts Texts mined from the Web Why collect texts? Leveraging existing translation solutions Creating a knowledge base, in some cases, an enterprise-wide knowledge base Leveraging existing terminology solutions
8	Terminology Extraction When is it most useful? If you have a good corpus available, particularly if you have parallel or aligned texts If your source text is in machine-readable form When wouldn’t I use it? When the source text is not in electronic form When resources are scattered through many very small texts
9	Term Extraction What are the advantages to using Term Extract or other extraction tools? Easier working interface Easy generation of starter term entries Easy access to concordances & selection of context references Easy exportation when it works!
10	Terminology Management Why store terminological information? Repeatability and reproducibility: Reuse of previously researched equivalents Harmonization of usage across a translation team Application for internationalization phase – reuse by technical writers within enterprises Interaction in translation memory environments
11	Term Level before Translation Term candidate extraction, either “manual” or automatic Terminology research using text corpora Web-based screen scraping Concordancing, etc. Listing single word (simple) terms Listing multiword terms
12	Term-level during Translation Manual term lookup in interactive word-processing/term database environment Automatic term lookup using translation memory and other CAT tools Concordancing for comparison of fuzzy matches Assured consistency, accuracy, & spelling
13	Term-Level after Translation Terminology consistency check Non-allowed terminology check
14	Sample Bitext (Aligned Segments)
15	Difference between Segmentation & Bitext TM: Segments are stored separately. Reassembly of segments to form original source or target text only possible if one of these texts is fed into TM. Bitext: Segments are identified in situ and the whole text is retained in memory. Individual segments are identified and presented dynamically as needed. (Transcorpora™)
16	Text Alignment Why align existing translations? Convert legacy translations for use in creating new translations Create structured resources for term extraction Share existing translations with translation teams in an interactive way
17	Segment Level before Translation New text segmentation Previous source-target text alignment Indexing “Clean-up” phase TM prepared for leveraging – reuse of existing translation segments Potential resource for terminology & text-based translation research
18	Segment Level during Translation So-called “pre-translation” phase Translation memory lookup Fuzzy matching Concordance information Machine translation (esp. example-based) Valuable for controlled language texts Valuable for “gisting”
19	Segment Level after Translation Missing segment detection Grammar check Format retention or regeneration
20	Translation Workbench Tools Trados/SDL http://www.trados.com/ http://www.translationzone.com/ http://www.sdl.com/ Star http://www.star-group.net/eng/software/sprachtech/transit.html DejaVu http://www.atril.com/
21	Translation Workbench Tools TermSeek http://www.termseekinc.com/ MultiCorpora/MultiTrans http://www.multicorpora.ca/index_e.html I-Term http://www.danterm.dk/iterm.htm
22	Tool Interaction Combined tool components: Word-processing (e.g., Word) or other text processing (TagEditor) Automatic term lookup (MultiTerm) Term in context (Concordancer) Translation memory Why do it? Increased efficiency, increased accuracy
23	Project Environments Coordinating project stages in team environments Storing and processing data The master program(s) need to be able to find all the files in order to interact. MultiTerm: work off your own directory on the LAN. Read-write problems if you constantly cycle a floppy or zip disk. Save data to your LAN folder.
24	Infrastructure Client-service provider environment T9n/L10n team environment Document creation & management system Terminology database Translation memory database Networking environment
25	Translation Workflow & Billing Workflow in medium to large translation and localization projects Project management software Workflow in content management, globalization management systems
26	Tools