Questionnaire Results

From MultilingualWeb-LT EC Project Wiki
Jump to: navigation, search
  • Repsonses have been anonymised
  • 'X' indicates a response that did not provide substantive new input
  • blank fields mean no reponse was provided
  • square brackets contain editor additions
  • P = problem
  • M = metadata
  • W = web tool

ID Problems Meta-data Web Technologies Further discussion Dublin workshop Assessment
R1 Y N
  • P1: Transparency of processes;
  • P2: Quality measures;
  • P3: Web apis
  • P4: Feedback, recycling, adaptation
  • P5: Exchange of linguistic/non linguistic resources
  • M1: cultural and social context
  • M2: provenance of source
  • M3: likelihood [?]
  • M4: timeframe
  • M5: self-validation (for LT)
  • M6: feedback trigger [?]
  • M7: process related data
  • W1: need effective and efficient mark up for automation
  • P1: addressed by requirements on process and provenance
  • P2: supported by quality related data categories
  • P3: Web service/APIs currently out of scope
  • P4: possible requirement on provenance tracking
  • P5: address by context and resource
  • M1: addresses by context
  • M2: addressed by provenance
  • M3: unclear
  • M4: addressed by process trigger
  • M5: addressed by confidence score?
  • M6: unclear
  • M7: addressed by process trigger
  • W1: addressed in general
R3 Y Y
  • P1: Identifying content to translate
  • P2: Optimising cost (TM reuse, minimise content to translate)
  • P3: Lack of context (links between strings)
  • P1: addressed by combination of translate, droprule, and localeSpecificInforation
  • P2: adddressed indirectly via provenance
  • P3: address by context, BUT link between strings needs to be addressed
  • M1: language
  • M2: domain
  • M3: terminology
W1: current tools satisfactory Y Y
  • M1: covered by ITS lang DC
  • M2: Covered by proposed domain DC
  • M3: Covered by ITS terminology DC
  • W1: general point
R6 X
  • M1: handling characters that should not change direction moving between ltr and rtl (see link)
  • W1: tool need to automate insertion, checking and updating/correction of language-related meta-data
  • W2: need ability to override automatic choices
  • M1: to considered under directionality
  • W1: general point, largely addressed
  • W2: to be considered as non functional requirement
  • P1: Extracting translatable strings;
  • Format conversion, e.g. XLIFF, wikitext
  • W1: exsiting standards poorly implements leading to proprietary variants
  • W2: web service/APIs needed to support robust SOA orchestration and integration
  • P1: addresses through translate, drop rule, localespecific context
  • W1: general point - how to reduce room for poor implementations
  • W2: currently out of scope
R8 Y Y
  • P1: CMS-TMS interop
  • M1: translate
  • M2: terminology
  • W1: CMS import/export functions must be improved
  • W2: choice in translation memory systems should be preserved
  • P1: general requirements
  • M1: addressed by translate
  • M2: addressed by terminology
  • W1: general requirement - how to reduce room for poor implementations
  • W2: general requirement
  • P1: Lack of context (WYSIWYG);
  • Tool complexity at transaltion checking phase
  • M1: translate
  • M2: terminology
  • W1: tool too complex and error prone, simpler, higher quality tools needed
  • W2: more WYSIWYG support
  • P1: need to consider under context
  • M1: addressed by translate
  • M2: addressed by terminology
  • W1: general requirement - how to reduce room for poor implementations
  • W2: need to consider under context
R11 X X
  • W1: simpler and more user friendly tools
  • W2: more open source and cheaper tools
  • W1: general requirement - need to consider simplicity, in scope?
  • W2: general requirement - probably not in scope?
  • P1: Translating content containing tags;
  • P2: Content spread across multiple logical units in soruce
  • M1: Terminology
  • M2: Directionality
  • W1: better integration of terminology and support for changes in terminology during translation process
  • P1: need to be considered
  • P2: needs to be considered
  • M1: addressed by terminology
  • M2: addressed by directionality
  • W1: change in terminology to be considered
R13 X
  • M1: meta-data is clearly understood (non functional requirement)
  • W1: need to support diverse language groups within countries
  • M1: to be considered in conduct of WG
  • W1: to consider under language information
  • P1: Entering and presenting special characters
  • P1: covered by unicode
  • P1: On the fly translations
  • P1: consider under process trigger
  • P1: Lack target content preview;
  • P2: Loss of context information; inconsistent segmentation and resulting TM loss;
  • P3: Lack association with (multiple) glossaries and multimedia content;
  • P4: Handling raw content in CAT tools with proper context
  • M1: links to external references and applicable rules, at translation unit level
  • M2: explicit markup of segementation beginning and end in response to unreliable segmentation rule implementations in CAT tools, especially in tagged text
  • M3: links between content and related multimedia content, ranging from source code, screen shots, pictures, videos (including audio tracks, and voice overs [and subtitles], to ensure consistency between these different
  • P1: need to consider under context
  • P2: need to consider under context, also could ITS markup result in TM loss. Is segmentation rule in scope?
  • P3: need to consider linked to multiple external references
  • P4: need to consider under context
  • M1: need to consider linked to multiple external refenences
  • M2: need to address in idvalue - should segeneration be included more explicitly
  • M3: need to consider linked to other media - is there some link to media fragment linkage here?
  • P1: ensure the meaning of a word in English is actually the same as in the translated language
  • M1: Ensuring that the translated document resembles the original, including language [style] and tense
  • M2: ensuring Technical Terminology is accurate,
  • W1: better understanding of Technical Communications and Technical Writing and Document Presentation
  • P1: address through quality assurance
  • M1: address through quality assurance
  • M2: need to consider link between terminology and quality assurance?
  • W1: general requirement
  • P1: lack of CMS-L10N technology causes translators to copy content from CMS, use CAT then paste back; copy HTML and paste back into CMS or use XML or XLIFF from CMS DB such that context is lost
  • P2: Some CMS have problems with unicode
  • M1: Meta-data about the visual and functional context where the translatable text operates,
  • M2: context data provided as common object, interaction methods (e.g. dropdown menu, or positions, e.g. footer)
  • M3: meta-data to allow translation tool to play/show/render the object, action etc involving translatable text
  • P1: not in scope?
  • M1: need to address in context
  • M2: need to address in context
  • M3: need to address n context and in general