- From: CVS User fsasaki <cvsmail@w3.org>
- Date: Thu, 05 Sep 2013 16:08:24 +0000
- To: public-multilingualweb-lt-commits@w3.org
Update of /w3ccvs/WWW/International/multilingualweb/lt/drafts/its20
In directory gil:/tmp/cvs-serv18183
Modified Files:
its20.html its20.odd
Log Message:
edits to resolve lqi issue-132
--- /w3ccvs/WWW/International/multilingualweb/lt/drafts/its20/its20.html 2013/09/05 15:47:58 1.491
+++ /w3ccvs/WWW/International/multilingualweb/lt/drafts/its20/its20.html 2013/09/05 16:08:23 1.492
@@ -142,9 +142,10 @@
<div class="toc1">E <a href="#informative-references" shape="rect">References</a> (Non-Normative)</div>
<div class="toc1">F <a href="#conversion-to-nif" shape="rect">Conversion to NIF</a> (Non-Normative)</div>
<div class="toc1">G <a href="#nif-backconversion" shape="rect">Conversion NIF2ITS</a> (Non-Normative)</div>
-<div class="toc1">H <a href="#list-of-elements-and-attributes" shape="rect">List of ITS 2.0 Global Elements and Local Attributes</a> (Non-Normative)</div>
-<div class="toc1">I <a href="#revisionlog" shape="rect">Revision Log</a> (Non-Normative)</div>
-<div class="toc1">J <a href="#acknowledgements" shape="rect">Acknowledgements</a> (Non-Normative)</div>
+<div class="toc1">H <a href="#localization-quality-guidance" shape="rect">Localization Quality Guidance</a> (Non-Normative)</div>
+<div class="toc1">I <a href="#list-of-elements-and-attributes" shape="rect">List of ITS 2.0 Global Elements and Local Attributes</a> (Non-Normative)</div>
+<div class="toc1">J <a href="#revisionlog" shape="rect">Revision Log</a> (Non-Normative)</div>
+<div class="toc1">K <a href="#acknowledgements" shape="rect">Acknowledgements</a> (Non-Normative)</div>
</div><hr/><div class="body"><div class="div1">
<h2><a href="#contents" shape="rect"><img src="images/topOfPage.gif" align="right" height="26" width="26" title="Go to the table of contents." alt="Go to the table of contents."/></a><a name="introduction" id="introduction" shape="rect"/>1 Introduction</h2><p>
<em>This section is informative.</em>
@@ -495,7 +496,7 @@
<strong class="hl-tag" style="color: #000096"></html></strong></pre></div><p>[Source file: <a href="examples/html5/EX-translate-html5-inline-global-1.html" shape="rect">examples/html5/EX-translate-html5-inline-global-1.html</a>]</p></div></div><div class="div3">
<h4><a href="#contents" shape="rect"><img src="images/topOfPage.gif" align="right" height="26" width="26" title="Go to the table of contents." alt="Go to the table of contents."/></a><a name="html5-its-local-markup" id="html5-its-local-markup" shape="rect"/>2.5.2 Local approach</h4><p>In HTML, an ITS 2.0 local data category is realized with the prefix <code>its-</code>.
The general mapping of the XML based ITS 2.0 attributes to their HTML counterparts is defined in
- <a class="section-ref" href="#html5-local-attributes" shape="rect">Section 6.1: Mapping of Local Data Categories to HTML</a>. An informative table in <a class="section-ref" href="#list-of-elements-and-attributes" shape="rect">Appendix H: List of ITS 2.0 Global Elements and Local Attributes</a>
+ <a class="section-ref" href="#html5-local-attributes" shape="rect">Section 6.1: Mapping of Local Data Categories to HTML</a>. An informative table in <a class="section-ref" href="#list-of-elements-and-attributes" shape="rect">Appendix I: List of ITS 2.0 Global Elements and Local Attributes</a>
provides an overview of the mapping for all data categories.</p></div><div class="div3">
<h4><a href="#contents" shape="rect"><img src="images/topOfPage.gif" align="right" height="26" width="26" title="Go to the table of contents." alt="Go to the table of contents."/></a><a name="html5-existing-markup-versus-its" id="html5-existing-markup-versus-its" shape="rect"/>2.5.3 HTML markup with ITS 2.0 counterparts</h4><p>There are four ITS 2.0 data categories, which have counterparts in HTML markup. In these cases, native HTML markup provides some information
in terms of ITS 2.0 data categories. For these data categories, ITS 2.0 defines the following:</p><ul><li><p>The <a href="#language-information" shape="rect">Language Information</a> data category has the HTML <code>lang</code>
@@ -1063,7 +1064,7 @@
generated those data category annotations.</p><div class="note"><p class="prefix"><b>Note:</b></p><ul><li><p id="annotators-ref-usage-scenarios">Three cases of providing tool
information can be expected:</p><ol class="depth1"><li><p>information about tools used for creating or modifying the textual
content;</p></li><li><p>information about tools that do 1), but also create ITS annotations, see
- <a class="section-ref" href="#list-of-elements-and-attributes" shape="rect">Appendix H: List of ITS 2.0 Global Elements and Local Attributes</a>; </p></li><li><p>information about tools that don’t modify or create content, but just
+ <a class="section-ref" href="#list-of-elements-and-attributes" shape="rect">Appendix I: List of ITS 2.0 Global Elements and Local Attributes</a>; </p></li><li><p>information about tools that don’t modify or create content, but just
create ITS annotations.</p></li></ol><p>
<code class="its-attr-markup">annotatorsRef</code> is only meant to be used when actual ITS
annotation is involved, that is for 2) and 3). To express tool information related
@@ -2603,8 +2604,8 @@
<h4><a href="#contents" shape="rect"><img src="images/topOfPage.gif" align="right" height="26" width="26" title="Go to the table of contents." alt="Go to the table of contents."/></a><a name="lqissue-definition" id="lqissue-definition" shape="rect"/>8.16.1 Definition</h4><p>The <a href="#lqissue" shape="rect">Localization Quality Issue</a> data category is used to
express information related to localization quality assessment tasks. Such tasks can
be conducted on the translation of some source content (such as a text or an image) into a target language or on the
- source content itself where its quality may impact on the localization process.</p><div class="note"><p class="prefix"><b>Note:</b></p><p>Automated or manual quality assessment is one area of quality management for translational and localization. An example of existing quality assessment is in-country review (e.g., as part of a language acceptance test for software). An important part of quality assessment is the list of issue types that are being used. Very often, simple issue categories like "correct/incorrect" or "like/dislike" are inadequate; instead, more specific ones such as "terminology" or "grammar" are more helpful in identifying concrete reasons for quality problems and for obtaining a more objective picture of quality levels.</p><p>Non-normative terminology related to localization quality as used in this section is provided in Appendix @@@TBD@@@.</p></div><p>This data category can be used in a number of ways, including the following example
- scenarios:</p><ul><li><p>A human reviewer working with a web-based tool adds quality markup manually in a text editor, including comments and suggestions, to localized content as part of the review process. A subsequent process examines this markup to ensure that changes were made.</p></li><li><p>A fully automatic quality checking tool flags a number of potential quality issues in an XML or HTML file and marks them up using ITS 2.0 markup. A human reviewer then uses another tool to examine this markup and decide whether the file needs to receive more extensive review or be passed on for further processing without a further manual review stage.</p></li><li><p>A quality assessment process identifies a number of issues and adds the ITS markup to a rendered HTML preview of an XML file along with CSS styling that highlights these issues. The resulting HTML file is then sent back to the translator to assist his or her revision efforts.</p></li></ul><div class="note"><p class="prefix"><b>Note:</b></><p id="selecting-issues">What issues should be considered in quality assessment tasks depends on the nature of the project and tools used. Further guidance is beyond the scope of this specification, but implementers may wish to consult the references cited in Appendix @@@TBD@@@.</p></div><p>The data category defines five pieces of information:</p><a name="lqissueDefs" id="lqissueDefs" shape="rect"/><table border="1" width="100%"><thead><tr><td rowspan="1" colspan="1">Information</td><td rowspan="1" colspan="1">Description</td><td rowspan="1" colspan="1">Value</td><td rowspan="1" colspan="1">Notes</td></tr></thead><tbody><tr><td rowspan="1" colspan="1">Type</td><td rowspan="1" colspan="1">A classifier that groups similar issues into categories (for example to differentiate spelling errors from grammar errors).</td><td rowspan="1" colspan="1">One of the values defined in <a href="#lqissue-typevalues" shape="rect">list of type
+ source content itself where its quality may impact on the localization process.</p><div class="note"><p class="prefix"><b>Note:</b></p><p>Automated or manual quality assessment is one area of quality management for translational and localization. An example of existing quality assessment is in-country review (e.g., as part of a language acceptance test for software). An important part of quality assessment is the list of issue types that are being used. Very often, simple issue categories like "correct/incorrect" or "like/dislike" are inadequate; instead, more specific ones such as "terminology" or "grammar" are more helpful in identifying concrete reasons for quality problems and for obtaining a more objective picture of quality levels.</p><p>Non-normative terminology related to localization quality as used in this section is provided in <a class="section-ref" href="#localization-quality-guidance" shape="rect">Appendix H: Localization Quality Guidance</a>.</p></div><p>This data category canbe used in a number of ways, including the following example
+ scenarios:</p><ul><li><p>A human reviewer working with a web-based tool adds quality markup manually in a text editor, including comments and suggestions, to localized content as part of the review process. A subsequent process examines this markup to ensure that changes were made.</p></li><li><p>A fully automatic quality checking tool flags a number of potential quality issues in an XML or HTML file and marks them up using ITS 2.0 markup. A human reviewer then uses another tool to examine this markup and decide whether the file needs to receive more extensive review or be passed on for further processing without a further manual review stage.</p></li><li><p>A quality assessment process identifies a number of issues and adds the ITS markup to a rendered HTML preview of an XML file along with CSS styling that highlights these issues. The resulting HTML file is then sent back to the translator to assist his or her revision efforts.</p></li></ul><div class="note"><p class="prefix"><b>Note:</b></><p id="selecting-issues">What issues should be considered in quality assessment tasks depends on the nature of the project and tools used. Further guidance is beyond the scope of this specification, but implementers may wish to consult the references cited in <a class="section-ref" href="#localization-quality-guidance" shape="rect">Appendix H: Localization Quality Guidance</a>.</p></div><p>The data category defines five pieces of information:</p><a name="lqissueDefs" id="lqissueDefs" shape="rect"/><table border="1" width="100%"><thead><tr><td rowspan="1" colspan="1">Information</td><td rowspan="1" colspan="1">Description</td><td rowspan="1" colspan="1">Value</td><td rowspan="1" colspan="1">Notes</td></tr></thead><tbody><tr><td rowspan="1" colspan="1">Type</td><td rowspan="1" colspan="1">A classifier that groups similar issues into categories (for example to differentiate spelling errors from grammar errors).</td><td rowspan="1" colspan="1">One of the values defined in <a href="#lqissue-typevalues" shape="ect">list of type
values</a>.</td><td rowspan="1" colspan="1">ITS 2.0-compliant tools that use these types <a href="#rfc-keywords" shape="rect">MUST</a> map their internal values to these types. If the type of the issue
is set to <code>uncategorized</code>, a comment <a href="#rfc-keywords" shape="rect">MUST</a> be specified as well.</td></tr><tr><td rowspan="1" colspan="1">Comment</td><td rowspan="1" colspan="1">A human-readable description of a specific instance of a quality issue.</td><td rowspan="1" colspan="1">Text</td><td rowspan="1" colspan="1">Comments can be used to explain an issue or provide guidance in addressing an issue. For example, a note about a Terminology issue might specify what term should be used.</td></tr><tr><td rowspan="1" colspan="1">Severity</td><td rowspan="1" colspan="1">A classifier for the seriousness of an issue. The seriousness depends on the Quality Model that is being applied. The Quality Model should be made explicit via the Profile Reference.</td><td rowspan="1" colspan="1">A rational number in the interval 0 to 100 (inclusive). The value follows the
<a href="http://www.w3.org/TR/2004/REC-xmlschema-2-20041028/#double" shape="rect">XML
@@ -3215,7 +3216,7 @@
quality issue types common in fully automatic and manual localization quality assessment. Tools
can map their internal types to these types in order to exchange information about the
kinds of issues they identify and take appropriate action even if another tool does not
- know the specific issues identified by the generating tool.</p><div class="note"><p class="prefix"><b>Note:</b></p><p>Note: The values of locQualityIssueType were derived from an early version of the QTLaunchPad project's Multidimensional Quality Metrics (MQM) framework. MQM is based on a careful analysis of existing translation quality assessment tools and models, such as the LISA QA Model, SAE J2450 @@@tbd - add reference?@@@, and various commercial tools. The values represent common issue types found in those models and are designed to provide interoperability between models. Differences in granularity and in issue types may prevent full interoperability, but using the shared values will maximize interoperability where possible.</p></div><p>The scope column in the following table identifies whether the issue type applies to the
+ know the specific issues identified by the generating tool.</p><div class="note"><p class="prefix"><b>Note:</b></p><p>Note: The values of locQualityIssueType were derived from an early version of the QTLaunchPad project's Multidimensional Quality Metrics (MQM) framework. MQM is based on a careful analysis of existing translation quality assessment tools and models, such as the LISA QA Model, SAE J2450, and various commercial tools. The values represent common issue types found in those models and are designed to provide interoperability between models. Differences in granularity and in issue types may prevent full interoperability, but using the shared values will maximize interoperability where possible.</p></div><p>The scope column in the following table identifies whether the issue type applies to the
source content (“S”), target content (“T”) or both (“S or T”).</p><p>The values listed in the following table are allowed for <code class="its-attr-markup">locQualityIssueType</code>.
Ideally the values a tool implementing the data category produces for the attribute matches one
of the values provided in this table and are as semantically accurate as possible. For example, marking the phrase “These man is” as a <code>terminology</code> issue, rather than as a <code>grammar</code> issue would be semantically inaccurate. Tools are encouraged to
@@ -3246,7 +3247,7 @@
<code>omission</code>
</td><td rowspan="1" colspan="1">Necessary text has been omitted from the localization or source.</td><td rowspan="1" colspan="1">
<ul><li><p>One or more segments found in the source that have been intended for translation are
- missing in the target.</p></li></ul>
+ missing in the target.</p></li><li><p>After an alignment, a verification tool flags the pairs of aligned segments where the target has no corresponding source because of incorrect segmentation or some alignment issue. In such case the 'omission' type may apply to the source entry.</p></li></ul>
</td><td rowspan="1" colspan="1">S or T</td><td rowspan="1" colspan="1">This value is not to be used for missing whitespace or formatting codes, but
instead has to be reserved for linguistic content.</td></tr><tr><td rowspan="1" colspan="1">
<code>untranslated</code>
@@ -3318,8 +3319,7 @@
<code>characters</code>
</td><td rowspan="1" colspan="1">The text contains characters that are garbled or incorrect or that are not used in
the language in which the content appears.</td><td rowspan="1" colspan="1">
- <ul><li><p>A text ought to have a '•' but instead has a '¥' sign.</p></li><li><p>A text translated into German systematically transforms 'ü', 'ö', and 'ä' to
- 'û', 'ô', and 'â'.</p></li><li><p>A Japanese text has been garbled and appears with Devanagari
+ <ul><li><p>A text ought to have a '•' but instead has a '¥' sign.</p></li><li><p>A German text erroneously uses û, ô, and â instead of the appropriate 'ü', 'ö', and 'ä'.</p></li><li><p>A Japanese text has been garbled and appears with Devanagari
characters.</p></li></ul>
</td><td rowspan="1" colspan="1">S or T</td><td rowspan="1" colspan="1">Characters ought to be used in cases of garbling or systematic use of
inappropriate characters, not for spelling issues where individual characters
@@ -3350,8 +3350,7 @@
</td><td rowspan="1" colspan="1">Numbers are inconsistent between source and target.</td><td rowspan="1" colspan="1">
<ul><li><p>A source text states that an object is 120 cm long, but the target text says
that it is 129 cm. long.</p></li></ul>
- </td><td rowspan="1" colspan="1">S or T</td><td rowspan="1" colspan="1">Some tools may correct for differences in units of measurement to reduce false
- positives.</td></tr><tr><td rowspan="1" colspan="1">
+ </td><td rowspan="1" colspan="1">S or T</td><td rowspan="1" colspan="1">Some tools may correct for differences in units of measurement to reduce false positives (e.g., a tool might adjust for differences in values between inches and centimeters to avoid flagging numbers that seem to be different but are in fact equivalent).</td></tr><tr><td rowspan="1" colspan="1">
<code>markup</code>
</td><td rowspan="1" colspan="1">There is an issue related to markup or a mismatch in markup between source and
target.</td><td rowspan="1" colspan="1">
@@ -3362,8 +3361,8 @@
that defines non-allowable content).</td><td rowspan="1" colspan="1">
<ul><li><p>The tool disallows the regular expression pattern ['"”’][\.,] but the
translated text contains "A leading “expert”, a political hack, claimed
- otherwise."</p></li></ul>
- </td><td rowspan="1" colspan="1">S or T</td><td rowspan="1" colspan="1"/></tr><tr><td rowspan="1" colspan="1">
+ otherwise."</p></li><li><p>A tool uses a regular expression to ensure that the content of an element is an IRI and flags what appears to be a malformed IRI.</p></li></ul>
+ </td><td rowspan="1" colspan="1">S or T</td><td rowspan="1" colspan="1">Defining what is or is not an allowable pattern is up to the processing application and is beyond the scope of this specification. Best practice would be to use the Comment attribute to specify the pattern that led to the issue.</td></tr><tr><td rowspan="1" colspan="1">
<code>whitespace</code>
</td><td rowspan="1" colspan="1">There is a mismatch in whitespace between source and target content or the text
violates specific rules related to the use of whitespace.</td><td rowspan="1" colspan="1">
@@ -3384,13 +3383,8 @@
</td><td rowspan="1" colspan="1">S or T</td><td rowspan="1" colspan="1">What constitutes a "significant" difference in length is determined by the model
referred to in the <code class="its-attr-markup">locQualityIssueProfileRef</code>.</td></tr><tr><td rowspan="1" colspan="1">
<code>non-conformance</code>
- </td><td rowspan="1" colspan="1">The content is deemed to have a level of conformance to a reference corpus. The
- <code>non-conformance</code> type reflects the degree to which the text conforms to
- a reference corpus given an algorithm, which combines several classes of error type to
- produce an aggregate rating. Higher values reflect poorer conformance.</td><td rowspan="1" colspan="1">The sentence "The harbour connected which to printer is busy or configared not
- properly." would have poor conformance.</td><td rowspan="1" colspan="1">S or T</td><td rowspan="1" colspan="1">In a system that uses classification techniques the poor conformance is a function
- of the combined incorrect terminology, wrong spelling and bad grammar, or other
- features as determined by the system.</td></tr><tr><td rowspan="1" colspan="1">
+ </td><td rowspan="1" colspan="1">The content is deemed to show poor statistical conformance to a reference corpus. Higher severity values reflect poorer conformance.</td><td rowspan="1" colspan="1">The sentence "The harbour connected which to printer is busy or configared not
+ properly." would have poor conformance.</td><td rowspan="1" colspan="1">S or T</td><td rowspan="1" colspan="1">Non-conformance is determined through the use of multiple statistical measures of similarity to a corpus of known-good content. For example, in a system that uses classification techniques the poor conformance might be a function of combined incorrect terminology, wrong spelling and bad grammar, or other features as determined by the system.</td></tr><tr><td rowspan="1" colspan="1">
<code>uncategorized</code>
</td><td rowspan="1" colspan="1">The issue either has not been categorized or cannot be categorized.</td><td rowspan="1" colspan="1">
<ul><li><p>A new version of a tool returns information on an issue that has not been
@@ -3417,7 +3411,7 @@
while other issues cannot.</p></li><li><p>If a system has an "miscellaneous" or "other" value, it is better to map this to this
value even if the specific instance of the issue might be mapped to another
value.</p></li></ul>
- </td></tr></tbody></table></div><div class="div1">
+ </td></tr></tbody></table><div class="note"><p class="prefix"><b>Note:</b></p><p>Note: <code>uncategorized</code> is used for issues that have not (yet) been categorized into a more specific value. For example, an automatic process might flag issues for attention but not provide any further detail or categorization: such issues would be listed as <code>uncategorized</code> in ITS 2.0. It may also be used when the exact nature of an issue is unclear and it cannot be categorized as a result (e.g., text is seriously garbled and the cause it unclear). By contrast other is used when the nature of an issue is clear but it cannot be categorized in one of the ITS 2.0 categories (or when a model or tool has its own “other” category). For example, in translation of subtitles there is a “respeaking” error category that does not correspond to any ITS 2.0 category and is highly specific to that environment; respeaking errors would therefore be categorized as <code>other</code> in ITS 2.0.</p><div></div><div class="div1">
<h2><a href="#contents" shape="rect"><img src="images/topOfPage.gif" align="right" height="26" width="26" title="Go to the table of contents." alt="Go to the table of contents."/></a><a name="its-schemas" id="its-schemas" shape="rect"/>D Schemas for ITS</h2><p>
<em>This section is normative.</em>
</p><p>The following schemas define ITS elements and attributes and can be used as building
@@ -5764,7 +5758,8 @@
</html></pre></div></div><p>Case 3: The NLP annotation created in NIF starts in one region and ends in another.
Solution: No straight mapping is possible; a mapping can be created if both regions have
the same parent.</p></div><div class="div1">
-<h2><a href="#contents" shape="rect"><img src="images/topOfPage.gif" align="right" height="26" width="26" title="Go to the table of contents." alt="Go to the table of contents."/></a><a name="list-of-elements-and-attributes" id="list-of-elements-and-attributes" shape="rect"/>H List of ITS 2.0 Global Elements and Local Attributes (Non-Normative)</h2><p>The following table lists global ITS 2.0 elements inside <code class="its-elem-markup">rules</code> element and local
+<h2><a href="#contents" shape="rect"><img src="images/topOfPage.gif" align="right" height="26" width="26" title="Go to the table of contents." alt="Go to the table of contents."/></a><a name="localization-quality-guidance" id="localization-quality-guidance" shape="rect"/>H Localization Quality Guidance (Non-Normative)</h2><p>The <a href="#lqissue" shape="rect">Localization Quality Issue</a> data category description uses the following terms as defined below for the purposes of this document.</p><ul><li><p><em>Quality assessment</em>. The task of evaluating the quality of translated content to determine its quality and to assign a value to it. Localization quality assessment is commonly conducted by identifying, categorizing, and counting issues in the translated content.</p></li><li><p><em>Issue</em>. A quality issue is a potential error detected in content. Issues may be detected automatically (e.g., by using a grammar checker or translation-specific tool) or manually, by human checking of content. Issuesmay or may not be errors (e.g., an apparent mistranslation may be deliberate and appropriate in some contexts) and should be confirmed by review.</p></li><li><p><em>Metric</em>. A metric is a formal system used in quality assessment tasks to identify issues, evaluate them, and determine quality. Metrics provide specific reference points for categorizing issues (as opposed to subjective assessment of quality, which does not use a metric) and may include weights for issues.</p></li><li><p><em>Model</em>. A model is the underlying description of the system that underlies a metric. (For example, some models may allow variable weights to be assigned to different issue types, in which case the specific metric used for a task will have these weights defined, even though the underlying model does not.)</p></li><li><p><em>Profile</em>. A quality profile is the adaptation of a model to specific requirements. It specifies specific conditions for using a model. It may include instructions and other guidelines that are ot included in the actual metric used. If a model allows for no customization, it has a single profile that is identical to the model; if it allows customization, each customization is a distinct profile.</p></li><li><p><em>Review</em>. The task of examining a text to identify any issues that occur in it. Review may be tied to the task of fixing any issues, a task generally referred to as revision.</p></li><li><p><em>Specifications</em>. Specifications (sometimes called a translation brief) are a description of the various expectations and requirements for a translation task. These may include statements about the type of translation expected, guidance on terminology to be used, information about audience, and so forth. Translation specifications are described in detail in ISO/TS-11669.</p></li><li><p><em>Tool</em>. As used here, a tool is software that generates localization quality markup. Tools may be fully automatic (e.g., a tool that identifies potential issues with terminology and grammar and marks thm without human intervention) or may required human input (e.g., a system that allows users to highlight spans of text and mark them with appropriate issues).</p></li></ul><p>For more information on setting translation project specifications and determining quality expectations, implementers are encouraged to consult the ISO standard definition of translation project specifications included in <a title="Translation projects – General guidance" href="#isots11669" shape="rect">[ISO/TS 11669:2002]</a>. Details about translation specifications are available at <a title="Structured Specifications and Translation Parameters" href="#structuredspecs" shape="rect">[Structured Specifications]</a>. While these documents do not directly address the definition of quality metrics, they provide useful guidance for implementers interested in determining which localization quality issue values should be used for specific scenarios.</p><p>The issue types defined in Localization Quality Issue were derived from the QTLaunchPd project’s Multidimensional Quality Metrics (MQM) framework. Additional guidance on this project may be found at [Multidimensional Quality Metrics].</p><p>The topic of localization quality is rapidly evolving and ITS 2.0 represents the first step in standardizing this area and will serve for basic interoperability needs. For situations requiring additional expressive capability or categories, further custom markup may be required.</p></div><div class="div1">
+<h2><a href="#contents" shape="rect"><img src="images/topOfPage.gif" align="right" height="26" width="26" title="Go to the table of contents." alt="Go to the table of contents."/></a><a name="list-of-elements-and-attributes" id="list-of-elements-and-attributes" shape="rect"/>I List of ITS 2.0 Global Elements and Local Attributes (Non-Normative)</h2><p>The following table lists global ITS 2.0 elements inside <code class="its-elem-markup">rules</code> element and local
ITS 2.0 markup in XML and HTML. Note that for the local markup there are various
constraints on what local attributes can be used together. Here these constraints are
expressed via occurrence indicators: optional "?", alternatives "|", or
@@ -5922,7 +5917,7 @@
<code class="its-attr-markup">lineBreakType</code>?</td><td rowspan="1" colspan="1">
<code class="its-attr-markup">its-storage-size</code>, <code class="its-attr-markup">its-storage-encoding</code>?,
<code class="its-attr-markup">lits-line-break-type</code>?</td></tr></tbody></table></div><div class="div1">
-<h2><a href="#contents" shape="rect"><img src="images/topOfPage.gif" align="right" height="26" width="26" title="Go to the table of contents." alt="Go to the table of contents."/></a><a name="revisionlog" id="revisionlog" shape="rect"/>I Revision Log (Non-Normative)</h2><p id="changelog-since-20130820">The following log records major changes that have been made to this document since the <a href="http://www.w3.org/TR/2013/WD-its20-20130820/" shape="rect">ITS 2.0 Working Draft 20 August 2013</a>:</p><ol class="depth1"><li><p>Updated reference to HTML5, see related <a href="http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2013Sep/0004.html" shape="rect">mail thread</a>.</p></li></ol><p id="changelog-since-20130521">The following log records major changes that have been made to this document since the <a href="http://www.w3.org/TR/2013/WD-its20-20130521/" shape="rect">ITS 2.0 Working Draft 21 May 2013</a>:</p><ol class="depth1"><li><p>Updated all text of the <a href="#html5-withintext-handling" sape="rect">HTML5 defaults for Element Within Text</a> and added example.
+<h2><a href="#contents" shape="rect"><img src="images/topOfPage.gif" align="right" height="26" width="26" title="Go to the table of contents." alt="Go to the table of contents."/></a><a name="revisionlog" id="revisionlog" shape="rect"/>J Revision Log (Non-Normative)</h2><p id="changelog-since-20130820">The following log records major changes that have been made to this document since the <a href="http://www.w3.org/TR/2013/WD-its20-20130820/" shape="rect">ITS 2.0 Working Draft 20 August 2013</a>:</p><ol class="depth1"><li><p>Updated reference to HTML5, see related <a href="http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2013Sep/0004.html" shape="rect">mail thread</a>.</p></li><li><p>Edits in <a class="section-ref" href="#lqissue" shape="rect">Section 8.16: Localization Quality Issue</a> and <a class="section-ref" href="#lqissue-typevalues" shape="rect">Appendix C: Values for the Localization Quality Issue Type</a> (plus a new informative <a class="section-ref" href="#localization-quality-guiance" shape="rect">Appendix H: Localization Quality Guidance</a>) to clarify the <a href="#lqissue" shape="rect">Localization Quality Issue</a> data category, see <a href="https://www.w3.org/International/multilingualweb/lt/track/issues/132" shape="rect">issue-132</a>.</p></li></ol><p id="changelog-since-20130521">The following log records major changes that have been made to this document since the <a href="http://www.w3.org/TR/2013/WD-its20-20130521/" shape="rect">ITS 2.0 Working Draft 21 May 2013</a>:</p><ol class="depth1"><li><p>Updated all text of the <a href="#html5-withintext-handling" shape="rect">HTML5 defaults for Element Within Text</a> and added example.
See <a href="https://www.w3.org/International/multilingualweb/lt/track/issues/118" shape="rect">issue-118</a>
and <a href="https://www.w3.org/International/multilingualweb/lt/track/actions/532" shape="rect">action-532</a>.</p></li><li><p>Added a paragraph about mime type submission to <a class="section-ref" href="#its-mime-type" shape="rect">Appendix B: Internationalization Tag Set (ITS) MIME Type</a>, see step 3 (first bullet point) at <a href="http://www.w3.org/2002/06/registering-mediatype#Planned" shape="rect">Register an Internet Media Type for a W3C Spec</a> and <a href="https://www.w3.org/International/multilingualweb/lt/track/actions/251" shape="rect">action-251</a>.</p></li><li><p>Removed company names from various examples, see <a href="http://www.w3.org/Search/Mail/Public/search?type-index=public-multilingualweb-lt-commits&index-type=t&keywords=action-502&search=Search" shape="rect">CVS commits</a> and <a href="https://www.w3.org/International/multilingualweb/lt/track/actions/502" shape="rect">action-502</a>.</p></li><li><p>Reformatting of various examples, see <a href="http://lsts.w3.org/Archives/Public/public-multilingualweb-lt-commits/2013May/0066.html" shape="rect">CVS commit mail</a> and further CVS commit mails with the same send time.</p></li><li><p>Edits related to <a title="" href="#nif-reference" shape="rect">[NIF]</a>, see <a href="https://www.w3.org/International/multilingualweb/lt/track/issues/125" shape="rect">issue-125</a>.</p></li><li><p>Edit related to <a href="https://www.w3.org/International/multilingualweb/lt/track/actions/527" shape="rect">action-527</a>: put the sentence about foreign elements / attributes also in <a class="section-ref" href="#datacategories-defaults-etc" shape="rect">Section 8.1: Position, Defaults, Inheritance, and Overriding of Data Categories</a>.</p></li><li><p>Copy editing of spec, see <a href="https://www.w3.org/International/multilingualweb/lt/track/actions/422" shape="rect">action-422</a>.</p></li><li><p>Edits related to <a href="https://www.w3.org/International/multilingualweb/lt/track/issues/126" shape="rect">issue-126</a> (Minor ssue with quality types listing).</p></li><li><p>Edits related to <a href="https://www.w3.org/International/multilingualweb/lt/track/issues/127" shape="rect">issue-127</a> (Clarifying HTML5 translate and global rules).</p></li><li><p>Edits related to <a href="https://www.w3.org/International/multilingualweb/lt/track/issues/128" shape="rect">issue-128</a> (Various editorial edits + conformance section fixes).</p></li><li><p>Edits related to <a href="https://www.w3.org/International/multilingualweb/lt/track/issues/129" shape="rect">issue-129</a> (sec1-2 editing (non-normative sec)).</p></li><li><p>Fixes about <a title="" href="#rfc2119" shape="rect">[RFC 2119]</a> statements, see <a href="https://www.w3.org/International/multilingualweb/lt/track/actions/540" shape="rect">action-540</a>.</p></li><li><p>Various checks: spelling (using U.S. English), style, grammar, use of ":" etc.</p></li><li><p>Added a non-normative XML Schema for ITS 2.0 to <a class="section-ref" href="#its-schemas" shape="rect">Appendix D: chemas for ITS</a>, see <a href="https://www.w3.org/International/multilingualweb/lt/track/actions/546" shape="rect">action-546</a>.</p></li><li><p>Updated link to quality issue type mappings in <a class="section-ref" href="#lqissue-typevalues" shape="rect">Appendix C: Values for the Localization Quality Issue Type</a>, see <a href="https://www.w3.org/International/multilingualweb/lt/track/actions/543" shape="rect">action-543</a>.</p></li><li><p>Clarify conflicts of <code class="its-elem-markup">param</code> elements with the same name in <a class="section-ref" href="#selection-precedence" shape="rect">Section 5.5: Precedence between Selections</a> and <a class="section-ref" href="#html5-selection-precedence" shape="rect">Section 6.4: Precedence between Selections</a>, see <a href="https://www.w3.org/International/multilingualweb/lt/track/issues/130" shape="rect">issue-130</a>.</p></li><li><p>Updated editors list.</p></li><li><p>Implemented <a href="http://lists.w3.org/Archives/Public/public-i18n-its-ig/013Jul/0041.html" shape="rect">editorial edits</a> in <a class="section-ref" href="#idvalue" shape="rect">Section 8.14: ID Value</a> and <a class="section-ref" href="#locNote-datacat" shape="rect">Section 8.3: Localization Note</a>.</p></li><li><p>Implemented <a href="http://lists.w3.org/Archives/Public/public-multilingualweb-lt-comments/2013Jul/0056.html" shape="rect">clarification</a> about <code class="its-attr-markup">annotatorsRef</code> attribute in <a class="section-ref" href="#its-tool-annotation" shape="rect">Section 5.7: ITS Tools Annotation</a>.</p></li><li><p>Implemented a <a href="http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2013Jul/0037.html" shape="rect">clarification</a> in <a class="section-ref" href="#mtconfidence-definition" shape="rect">Section 8.18.1: Definition</a> and added <a href="#mt-confidence-score-generation-tools" shape="rect">a note</a> that <a href="#mtconfidence" shape="rect">MT Confidence</a> information can be produced both by MT systems and other tool.</p></li><li><p>Changed the <a href="#conversion-to-nif" shape="rect">conversion to NIF</a> to be a non-normative feature.</p></li></ol><p id="changelog-since-20130411">The following log records major changes that have been made to this document since the <a href="http://www.w3.org/TR/2013/WD-its20-20130411/" shape="rect">ITS 2.0 Working Draft 11 April 2013</a>:</p><ol class="depth1"><li><p>Added a reference to the <a href="http://www.w3.org/2005/11/its/rdf#" shape="rect">ITS RDF Ontology</a> and an <a href="#its-rdf-ontology-status" shape="rect">explanatory note</a> about its status to <a class="section-ref" href="#conversion-to-nif" shape="rect">Appendix F: Conversion to NIF</a>, see <a href="https://www.w3.org/International/multilingualweb/lt/track/actions/514" shape="rect">action-514</a>.</p></li><li><p>Updated <a class="section-ref" href="#conversion-to-nif" shape="rect">Appendix F: Conversion to NIF</a> to reflect <a href="http://www.w3.org/2013/05/08-mlw-lt-minutes#item04" shape="rect">MLW-LT May 013 f2f discussion</a>: <code>nif:occursIn</code> has changed to <code>nif:sourceUrl</code>, and <code>nif:convertedFrom</code> replaces <code>itsrdf:xpath2nif</code>. See <a href="https://www.w3.org/International/multilingualweb/lt/track/actions/517" shape="rect">action-517</a>.</p></li><li><p>Added <a href="#local-approach-not-applicable-to-attributes" shape="rect">a note</a> to <a class="section-ref" href="#basic-concepts-selection-local" shape="rect">Section 2.2.1: Local Approach</a> expressing that local selection does not apply to attributes,
see <a href="https://www.w3.org/International/multilingualweb/lt/track/issues/98" shape="rect">issue-98</a>.</p></li><li><p>Added a <a href="#qa-issue-types-mappings" shape="rect">clarification</a> about the role of mappings from tools to quality issue types,
@@ -5976,7 +5971,7 @@
minutes</a>.</p></li><li><p>Made changes (see <a href="http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2012Nov/0266.html" shape="rect">detailed description</a>) to <a href="#lqissue-typevalues" shape="rect">descriptions of
allowed values</a> for <a href="#lqissue" shape="rect">Localization Quality Issue</a>
(specifically <em>terminology</em>, <em>locale-violation</em>, and
- <em>whitespace</em> to respond to and clarify <a href="http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2012Nov/0208.html" shape="rect">points raised by Daniel Naber</a>.</p></li><li><p>Added <a class="section-ref" href="#list-of-elements-and-attributes" shape="rect">Appendix H: List of ITS 2.0 Global Elements and Local Attributes</a>, see <a href="https://www.w3.org/International/multilingualweb/lt/track/actions/321" shape="rect">action-321</a>.</p></li><li><p>Renaming attribute for <a class="section-ref" href="#its-tool-annotation" shape="rect">Section 5.7: ITS Tools Annotation</a>. See <a href="http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2012Dec/0006.html" shape="rect">change description</a>.</p></li><li><p>Changes related to <code class="its-attr-markup">annotatorsRef</code>, see <a href="http://www.w3.org/2012/12/03-mlw-lt-minutes.html#item08" shape="rect">Working Group call
+ <em>whitespace</em> to respond to and clarify <a href="http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2012Nov/0208.html" shape="rect">points raised by Daniel Naber</a>.</p></li><li><p>Added <a class="section-ref" href="#list-of-elements-and-attributes" shape="rect">Appendix I: List of ITS 2.0 Global Elements and Local Attributes</a>, see <a href="https://www.w3.org/International/multilingualweb/lt/track/actions/321" shape="rect">action-321</a>.</p></li><li><p>Renaming attribute for <a class="section-ref" href="#its-tool-annotation" shape="rect">Section 5.7: ITS Tools Annotation</a>. See <a href="http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2012Dec/0006.html" shape="rect">change description</a>.</p></li><li><p>Changes related to <code class="its-attr-markup">annotatorsRef</code>, see <a href="http://www.w3.org/2012/12/03-mlw-lt-minutes.html#item08" shape="rect">Working Group call
2012-12-03</a> discussion.</p></li><li><p>Changes related to <code>disambigGranularity</code> attribute, see <a href="http://www.w3.org/2012/12/03-mlw-lt-minutes.html#item04" shape="rect">Working Group call
2012-12-03</a> discussion and <a href="https://www.w3.org/International/multilingualweb/lt/track/actions/359" shape="rect">action-359</a>.</p></li></ol><p id="changelog-since-20120829">The following log records major changes that have been
made to this document since the <a href="http://www.w3.org/TR/2012/WD-its20-20120829/" shape="rect">ITS 2.0 Working Draft 29 August 2012</a>:</p><ol class="depth1"><li><p>Added a first draft of <a class="section-ref" href="#provenance" shape="rect">Section 8.11: Provenance</a></p></li><li><p>Added <a class="section-ref" href="#html5-markup" shape="rect">Section 6: Using ITS Markup in HTML</a>.</p></li><li><p>Removed inline markup declarations.</p></li><li><p>Addition of a <code class="its-attr-markup">locQualityRatingVote</code> attribute and a
@@ -5998,6 +5993,6 @@
this document:</p><ol class="depth1"><li><p>Clarified <a href="#introduction" shape="rect">introduction</a> to cover ITS 2.0</p></li><li><p>Added a subsection on the relation to ITS 1.0 to the introduction.</p></li><li><p>Created HTML based declarations for various data categories, see e.g., HTML
declarations for the Terminology data category and the summary for local data categories
in <a class="section-ref" href="#selection-local" shape="rect">Section 5.2.2: Local Selection in an XML Document</a></p></li><li><p>Created examples for these declarations, see e.g., <a href="#EX-term-local-html-1" shape="rect">Example 40</a></p></li><li><p>Added placeholders for new data categories to <a class="section-ref" href="#datacategory-description" shape="rect">Section 8: Description of Data Categories</a></p></li><li><p>Added a placeholder section <a class="section-ref" href="#conversion-to-nif" shape="rect">Appendix F: Conversion to NIF</a></p></li></ol></div><div class="div1">
-<h2><a href="#contents" shape="rect"><img src="images/topOfPage.gif" align="right" height="26" width="26" title="Go to the table of contents." alt="Go to the table of contents."/></a><a name="acknowledgements" id="acknowledgements" shape="rect"/>J Acknowledgements (Non-Normative)</h2><p>This document has been developed with contributions by the
+<h2><a href="#contents" shape="rect"><img src="images/topOfPage.gif" align="right" height="26" width="26" title="Go to the table of contents." alt="Go to the table of contents."/></a><a name="acknowledgements" id="acknowledgements" shape="rect"/>K Acknowledgements (Non-Normative)</h2><p>This document has been developed with contributions by the
<a href="http://www.w3.org/International/multilingualweb/lt/" shape="rect">MultilingualWeb-LT Working Group</a> and collaborators: Mihael Arcan (DERI Galway at the National University of Ireland, Galway, Ireland), Pablo Badía (Linguaserve), Aaron Beaton (Opera Software), Renat Bikmatov (Logrus Plus LLC), Aljoscha Burchardt (German Research Center for Artificial Intelligence (DFKI GmbH)), Nicoletta CalzolarI (CNR--Consiglio Nazionale delle Ricerche), Somnath Chandra (Department of Information Technology, Government of India), John Colosi (Verisign, Inc.), Mauricio del Olmo (Linguaserve), Giuseppe Deriard (Linguaserve), Pedro Luis Díez Orzas (Linguaserve), David Filip (University of Limerick), Leroy Finn (Trinity College Dublin), Karl Fritsche (Cocomore AG), Serge Gladkoff (Logrus Plus LLC), Tatiana Gornostay (Tilde), Daniel Grasmick (Lucy Software and Services GmbH), Declan Groves (Centre for Next Generation Localisation), Manuel Honegger (University of Limerick), Dominic Jones (Trinity CollegeDublin), Matthias Kandora (]init[), Milan Karásek (Moravia Worldwide), Jirka Kosek (University of Economics, Prague), Michael Kruppa (Cocomore AG), Alejandro Leiva (Cocomore AG), Swaran Lata (Department of Information Technology, Government of India), David Lewis (Trinity College Dublin), Fredrik Liden (ENLASO Corporation), Christian Lieske (SAP AG), Qun Liu (Centre for Next Generation Localisation), Arle Lommel (German Research Center for Artificial Intelligence (DFKI GmbH)), Priyanka Malik (Department of Information Technology, Government of India), Shaun McCance ((public) Invited expert), Sean Mooney (University of Limerick), Jan Nelson (Microsoft Corporation), Pablo Nieto Caride (Linguaserve), Pēteris Ņikiforovs (Tilde), Naoto Nishio (University of Limerick), Philip O'Duffy (University of Limerick), Des Oates (Adobe Systems Inc.), Georgios Petasis (Institute of Informatics & Telecommunications (IIT), NCSR), Mārcis Pinnis (Tilde), Prashant Verma Prashant (Department of Information Technology, Govrnment of India), Georg Rehm (German Research Center for Artificial Intelligence (DFKI GmbH)), Phil Ritchie (VistaTEC), Thomas Rüdesheim (Lucy Software and Services GmbH), Nieves Sande (German Research Center for Artificial Intelligence (DFKI GmbH)), Felix Sasaki (DFKI / W3C Fellow), Yves Savourel (ENLASO Corporation), Jörg Schütz (W3C Invited Experts), Sebastian Sklarß (]init[), Ankit Srivastava (Centre for Next Generation Localisation), Tadej Štajner (Jozef Stefan Institute), Olaf-Michael Stefanov ((public) Invited expert), Najib Tounsi (Ecole Mohammadia d'Ingenieurs Rabat (EMI)), Naitik Tyagi Tyagi (Department of Information Technology, Government of India), Ronny Unger (Cocomore AG), Clemens Weins (Cocomore AG).</p><p>A special thanks goes to the following persons:
</p><ul><li><p>Sebastian Hellmann for introducing us to <a title="" href="#nif-reference" shape="rect">[NIF]</a> and for contributing to the creation of the <a href="http://www.w3.org/2005/11/its/rdf#" shape="rect">ITS 2.0 ontology</a> and NIF testing.</p></li><li><p>Daniel Naber for introducing us to <a href="http://languagetool.org" shape="rect">LanguageTool</a> and for implementing <a href="#lqissue-typevalues" shape="rect">Localization Quality Issue Type</a> functionality in language tool.</p></li></ul></div></div></body></html>
\ No newline at end of file
--- /w3ccvs/WWW/International/multilingualweb/lt/drafts/its20/its20.odd 2013/09/05 15:47:59 1.506
+++ /w3ccvs/WWW/International/multilingualweb/lt/drafts/its20/its20.odd 2013/09/05 16:08:24 1.507
@@ -3930,7 +3930,7 @@
source content itself where its quality may impact on the localization process.</p>
<note>
<p>Automated or manual quality assessment is one area of quality management for translational and localization. An example of existing quality assessment is in-country review (e.g., as part of a language acceptance test for software). An important part of quality assessment is the list of issue types that are being used. Very often, simple issue categories like "correct/incorrect" or "like/dislike" are inadequate; instead, more specific ones such as "terminology" or "grammar" are more helpful in identifying concrete reasons for quality problems and for obtaining a more objective picture of quality levels.</p>
- <p>Non-normative terminology related to localization quality as used in this section is provided in Appendix @@@TBD@@@.</p>
+ <p>Non-normative terminology related to localization quality as used in this section is provided in <ptr target="#localization-quality-guidance" type="specref"/>.</p>
</note>
<p>This data category can be used in a number of ways, including the following example
scenarios:</p>
@@ -3939,7 +3939,7 @@
<item>A fully automatic quality checking tool flags a number of potential quality issues in an XML or HTML file and marks them up using ITS 2.0 markup. A human reviewer then uses another tool to examine this markup and decide whether the file needs to receive more extensive review or be passed on for further processing without a further manual review stage.</item>
<item>A quality assessment process identifies a number of issues and adds the ITS markup to a rendered HTML preview of an XML file along with CSS styling that highlights these issues. The resulting HTML file is then sent back to the translator to assist his or her revision efforts.</item>
</list>
- <note><p xml:id="selecting-issues">What issues should be considered in quality assessment tasks depends on the nature of the project and tools used. Further guidance is beyond the scope of this specification, but implementers may wish to consult the references cited in Appendix @@@TBD@@@.</p></note>
+ <note><p xml:id="selecting-issues">What issues should be considered in quality assessment tasks depends on the nature of the project and tools used. Further guidance is beyond the scope of this specification, but implementers may wish to consult the references cited in <ptr target="#localization-quality-guidance" type="specref"/>.</p></note>
<p>The data category defines five pieces of information:</p>
<table border="1" xml:id="lqissueDefs" width="100%">
<row role="head">
@@ -4864,7 +4864,7 @@
can map their internal types to these types in order to exchange information about the
kinds of issues they identify and take appropriate action even if another tool does not
know the specific issues identified by the generating tool.</p>
- <note><p>Note: The values of locQualityIssueType were derived from an early version of the QTLaunchPad project's Multidimensional Quality Metrics (MQM) framework. MQM is based on a careful analysis of existing translation quality assessment tools and models, such as the LISA QA Model, SAE J2450 @@@tbd - add reference?@@@, and various commercial tools. The values represent common issue types found in those models and are designed to provide interoperability between models. Differences in granularity and in issue types may prevent full interoperability, but using the shared values will maximize interoperability where possible.</p></note>
+ <note><p>Note: The values of locQualityIssueType were derived from an early version of the QTLaunchPad project's Multidimensional Quality Metrics (MQM) framework. MQM is based on a careful analysis of existing translation quality assessment tools and models, such as the LISA QA Model, SAE J2450, and various commercial tools. The values represent common issue types found in those models and are designed to provide interoperability between models. Differences in granularity and in issue types may prevent full interoperability, but using the shared values will maximize interoperability where possible.</p></note>
<p>The scope column in the following table identifies whether the issue type applies to the
source content (“S”), target content (“T”) or both (“S or T”).</p>
<p>The values listed in the following table are allowed for <att>locQualityIssueType</att>.
@@ -4929,6 +4929,7 @@
<list>
<item>One or more segments found in the source that have been intended for translation are
missing in the target.</item>
+ <item>After an alignment, a verification tool flags the pairs of aligned segments where the target has no corresponding source because of incorrect segmentation or some alignment issue. In such case the 'omission' type may apply to the source entry.</item>
</list>
</cell>
<cell>S or T</cell>
@@ -5081,8 +5082,7 @@
<cell>
<list>
<item>A text ought to have a '•' but instead has a '¥' sign.</item>
- <item>A text translated into German systematically transforms 'ü', 'ö', and 'ä' to
- 'û', 'ô', and 'â'.</item>
+ <item>A German text erroneously uses û, ô, and â instead of the appropriate 'ü', 'ö', and 'ä'.</item>
<item>A Japanese text has been garbled and appears with Devanagari
characters.</item>
</list>
@@ -5154,8 +5154,7 @@
</list>
</cell>
<cell>S or T</cell>
- <cell>Some tools may correct for differences in units of measurement to reduce false
- positives.</cell>
+ <cell>Some tools may correct for differences in units of measurement to reduce false positives (e.g., a tool might adjust for differences in values between inches and centimeters to avoid flagging numbers that seem to be different but are in fact equivalent).</cell>
</row>
<row>
<cell><code>markup</code></cell>
@@ -5179,10 +5178,11 @@
<item>The tool disallows the regular expression pattern ['"”’][\.,] but the
translated text contains "A leading “expert”, a political hack, claimed
otherwise."</item>
+ <item>A tool uses a regular expression to ensure that the content of an element is an IRI and flags what appears to be a malformed IRI.</item>
</list>
</cell>
<cell>S or T</cell>
- <cell/>
+ <cell>Defining what is or is not an allowable pattern is up to the processing application and is beyond the scope of this specification. Best practice would be to use the Comment attribute to specify the pattern that led to the issue.</cell>
</row>
<row>
<cell><code>whitespace</code></cell>
@@ -5230,16 +5230,11 @@
</row>
<row>
<cell><code>non-conformance</code></cell>
- <cell>The content is deemed to have a level of conformance to a reference corpus. The
- <code>non-conformance</code> type reflects the degree to which the text conforms to
- a reference corpus given an algorithm, which combines several classes of error type to
- produce an aggregate rating. Higher values reflect poorer conformance.</cell>
+ <cell>The content is deemed to show poor statistical conformance to a reference corpus. Higher severity values reflect poorer conformance.</cell>
<cell>The sentence "The harbour connected which to printer is busy or configared not
properly." would have poor conformance.</cell>
<cell>S or T</cell>
- <cell>In a system that uses classification techniques the poor conformance is a function
- of the combined incorrect terminology, wrong spelling and bad grammar, or other
- features as determined by the system.</cell>
+ <cell>Non-conformance is determined through the use of multiple statistical measures of similarity to a corpus of known-good content. For example, in a system that uses classification techniques the poor conformance might be a function of combined incorrect terminology, wrong spelling and bad grammar, or other features as determined by the system.</cell>
</row>
<row>
<cell><code>uncategorized</code></cell>
@@ -5290,6 +5285,7 @@
</cell>
</row>
</table>
+ <note><p>Note: <code>uncategorized</code> is used for issues that have not (yet) been categorized into a more specific value. For example, an automatic process might flag issues for attention but not provide any further detail or categorization: such issues would be listed as <code>uncategorized</code> in ITS 2.0. It may also be used when the exact nature of an issue is unclear and it cannot be categorized as a result (e.g., text is seriously garbled and the cause it unclear). By contrast other is used when the nature of an issue is clear but it cannot be categorized in one of the ITS 2.0 categories (or when a model or tool has its own “other” category). For example, in translation of subtitles there is a “respeaking” error category that does not correspond to any ITS 2.0 category and is highly specific to that environment; respeaking errors would therefore be categorized as <code>other</code> in ITS 2.0.</p></note>
</div>
<div xml:id="its-schemas">
<head>Schemas for ITS</head>
@@ -5758,6 +5754,21 @@
Solution: No straight mapping is possible; a mapping can be created if both regions have
the same parent.</p>
</div>
+ <div xml:id="localization-quality-guidance" type="inform">
+ <head>Localization Quality Guidance</head>
+ <p>The <ref target="#lqissue">Localization Quality Issue</ref> data category description uses the following terms as defined below for the purposes of this document.</p>
+ <list type="unordered"><item><emph>Quality assessment</emph>. The task of evaluating the quality of translated content to determine its quality and to assign a value to it. Localization quality assessment is commonly conducted by identifying, categorizing, and counting issues in the translated content.</item>
+ <item><emph>Issue</emph>. A quality issue is a potential error detected in content. Issues may be detected automatically (e.g., by using a grammar checker or translation-specific tool) or manually, by human checking of content. Issues may or may not be errors (e.g., an apparent mistranslation may be deliberate and appropriate in some contexts) and should be confirmed by review.</item>
+ <item><emph>Metric</emph>. A metric is a formal system used in quality assessment tasks to identify issues, evaluate them, and determine quality. Metrics provide specific reference points for categorizing issues (as opposed to subjective assessment of quality, which does not use a metric) and may include weights for issues.</item>
+ <item><emph>Model</emph>. A model is the underlying description of the system that underlies a metric. (For example, some models may allow variable weights to be assigned to different issue types, in which case the specific metric used for a task will have these weights defined, even though the underlying model does not.)</item>
+ <item><emph>Profile</emph>. A quality profile is the adaptation of a model to specific requirements. It specifies specific conditions for using a model. It may include instructions and other guidelines that are not included in the actual metric used. If a model allows for no customization, it has a single profile that is identical to the model; if it allows customization, each customization is a distinct profile.</item>
+ <item><emph>Review</emph>. The task of examining a text to identify any issues that occur in it. Review may be tied to the task of fixing any issues, a task generally referred to as revision.</item>
+ <item><emph>Specifications</emph>. Specifications (sometimes called a translation brief) are a description of the various expectations and requirements for a translation task. These may include statements about the type of translation expected, guidance on terminology to be used, information about audience, and so forth. Translation specifications are described in detail in ISO/TS-11669.</item>
+ <item><emph>Tool</emph>. As used here, a tool is software that generates localization quality markup. Tools may be fully automatic (e.g., a tool that identifies potential issues with terminology and grammar and marks them without human intervention) or may required human input (e.g., a system that allows users to highlight spans of text and mark them with appropriate issues).</item></list>
+ <p>For more information on setting translation project specifications and determining quality expectations, implementers are encouraged to consult the ISO standard definition of translation project specifications included in <ptr target="#isots11669" type="bibref"/>. Details about translation specifications are available at <ptr target="#structuredspecs" type="bibref"/>. While these documents do not directly address the definition of quality metrics, they provide useful guidance for implementers interested in determining which localization quality issue values should be used for specific scenarios.</p>
+ <p>The issue types defined in Localization Quality Issue were derived from the QTLaunchPad project’s Multidimensional Quality Metrics (MQM) framework. Additional guidance on this project may be found at [Multidimensional Quality Metrics].</p>
+ <p>The topic of localization quality is rapidly evolving and ITS 2.0 represents the first step in standardizing this area and will serve for basic interoperability needs. For situations requiring additional expressive capability or categories, further custom markup may be required.</p>
+ </div>
<div xml:id="list-of-elements-and-attributes" type="inform">
<head>List of ITS 2.0 Global Elements and Local Attributes</head>
<p>The following table lists global ITS 2.0 elements inside <gi>rules</gi> element and local
@@ -5932,6 +5943,7 @@
<p xml:id="changelog-since-20130820">The following log records major changes that have been made to this document since the <ref target="http://www.w3.org/TR/2013/WD-its20-20130820/">ITS 2.0 Working Draft 20 August 2013</ref>:</p>
<list type="ordered">
<item>Updated reference to HTML5, see related <ref target="http://lists.w3.org/Archives/Public/public-multilingualweb-lt/2013Sep/0004.html">mail thread</ref>.</item>
+ <item>Edits in <ptr target="#lqissue" type="specref"/> and <ptr target="#lqissue-typevalues" type="specref"/> (plus a new informative <ptr target="#localization-quality-guidance" type="specref"/>) to clarify the <ref target="#lqissue">Localization Quality Issue</ref> data category, see <ref target="https://www.w3.org/International/multilingualweb/lt/track/issues/132">issue-132</ref>.</item>
</list>
<p xml:id="changelog-since-20130521">The following log records major changes that have been made to this document since the <ref target="http://www.w3.org/TR/2013/WD-its20-20130521/">ITS 2.0 Working Draft 21 May 2013</ref>:</p>
<list type="ordered">
Received on Thursday, 5 September 2013 16:08:31 UTC