GALE Chinese-English Word Alignment and Tagging Training Part 3 -- Web

<h3>Introduction</h3> <p>GALE Chinese-English Word Alignment and Tagging Training Part 3 -- Web was developed by the Linguistic Data Consortium (LDC) and contains 154,541 tokens of word aligned Chinese and English parallel text enriched with linguistic tags. This material was used as training data in the <a href="http://projects.ldc.upenn.edu/gale/index.html" rel="nofollow">DARPA GALE</a> (Global Autonomous Language Exploitation) program.</p> <p> Some approaches to statistical machine translation include the incorporation of linguistic knowledge in word aligned text as a means to improve automatic word alignment and machine translation quality. This is accomplished with two annotation schemes: alignment and tagging. Alignment identifies minimum translation units and translation relations by using minimum-match and attachment annotation approaches. A set of word tags and alignment link tags are designed in the tagging scheme to describe these translation units and relations. Tagging adds contextual, syntactic and language-specific features to the alignment annotation. </p> <p>GALE Chinese-English Word Alignment and Tagging Training Part 1 -- Newswire and Web (<a href="http://catalog.ldc.upenn.edu/LDC2012T16" rel="nofollow">LDC2012T16</a>) and GALE Chinese-English Word Alignment and Tagging Training Part 2 -- Newswire (<a href="http://catalog.ldc.upenn.edu/LDC2012T20" rel="nofollow">LDC2012T20</a>) are also available through LDC.</p> <h3>Data</h3> <p>This release consists of Chinese source web data (newsgroup, weblog) collected by LDC in 2008 and 2009. The distribution by words, character tokens and segments appears below: </p> <table> <tr> <td>Language</td> <td>Files</td> <td>Words</td> <td>CharTokens</td> <td>Segments</td> </tr> <tr> <td>Chinese</td> <td>1249</td> <td>103027</td> <td>154541</td> <td>4842</td> </tr> </table><p>Note that all token counts are based on the Chinese data only. One token is equivalent to one character and one word is equivalent to 1.5 characters.</p> <p> The Chinese word alignment tasks consisted of the following components: </p> <ul> <li>Identifying, aligning, and tagging 8 different types of links</li> <li>Identifying, attaching, and tagging local-level unmatched words</li> <li>Identifying and tagging sentence/discourse-level unmatched words</li> <li>Identifying and tagging all instances of Chinese 的 (DE) except when they were a part of a semantic link.</li> </ul><h3>Samples</h3> <ul> <li><a href="./desc/addenda/LDC2012T24.cmn.raw.txt" rel="nofollow">Chinese raw source sample</a></li> <li><a href="./desc/addenda/LDC2012T24.cmn.tkn.txt" rel="nofollow">Chinese character tokenized sample</a></li> <li><a href="./desc/addenda/LDC2012T24.eng.raw.txt" rel="nofollow">English raw translation sample</a></li> <li><a href="./desc/addenda/LDC2012T24.eng.tkn.txt" rel="nofollow">English tokenized sample</a></li> <li><a href="./desc/addenda/LDC2012T24.wa.txt" rel="nofollow">Character-based word alignment sample</a></li> </ul><h3>Sponsorship</h3> <p>This work was supported in part by the Defense Advanced Research Projects Agency, GALE Program Grant No. HR0011-06-1-0003. The content of this publication does not necessarily reflect the position or the policy of the Government, and no official endorsement should be inferred.</p> <h3>Updates</h3> <p> None at this time. </p> </br>
Portions © 2008, 2009, 2012 Trustees of the University of Pennsylvania

<h3>简介</h3> <p>GALE汉英词对齐与标注训练语料第三部分——网页版由语言数据联盟（Linguistic Data Consortium, LDC）开发，包含154541个词元（Token）的汉英平行对齐文本，且附带丰富的语言标注。该语料曾作为训练数据应用于<a href="http://projects.ldc.upenn.edu/gale/index.html" rel="nofollow">美国国防高级研究计划局（Defense Advanced Research Projects Agency, DARPA）的GALE（全球自主语言开发，Global Autonomous Language Exploitation）</a>项目。</p> <p>统计机器翻译的诸多研究思路中，会将语言知识融入词对齐文本，以此优化自动词对齐效果与机器翻译质量。这一目标可通过两种标注方案实现：词对齐与词标注。词对齐方案采用最小匹配与依附标注方法，识别最小翻译单元与翻译对应关系；标注方案则设计了一套词标签与对齐链接标签，用于描述上述翻译单元与对应关系，为对齐标注补充上下文、句法及语种专属特征。</p> <p>GALE汉英词对齐与标注训练语料第一部分——新闻专线与网页版（编号<a href="http://catalog.ldc.upenn.edu/LDC2012T16" rel="nofollow">LDC2012T16</a>）以及第二部分——新闻专线版（编号<a href="http://catalog.ldc.upenn.edu/LDC2012T20" rel="nofollow">LDC2012T20</a>）均可通过LDC获取。</p> <h3>数据详情</h3> <p>本次发布的数据为LDC于2008年与2009年采集的中文网页源文本（涵盖新闻组、博客内容）。文本的词数、字符词元（CharTokens）数与语段划分分布如下表所示：</p> <table> <tr> <td>语种</td> <td>文件数</td> <td>词数</td> <td>字符词元数</td> <td>语段数</td> </tr> <tr> <td>中文</td> <td>1249</td> <td>103027</td> <td>154541</td> <td>4842</td> </tr> </table><p>请注意，所有词元计数仅针对中文文本。1个词元等价于1个汉字，1个词等价于1.5个汉字。</p> <p>中文词对齐任务包含以下组成部分：</p> <ul> <li>识别、对齐并标注8种不同类型的链接</li> <li>识别、依附并标注局部级未匹配词</li> <li>识别并标注句子/语篇级未匹配词</li> <li>识别并标注所有中文“的”（DE）实例，语义链接中的“的”除外</li> </ul><h3>示例样本</h3> <ul> <li><a href="./desc/addenda/LDC2012T24.cmn.raw.txt" rel="nofollow">中文原始源文本示例</a></li> <li><a href="./desc/addenda/LDC2012T24.cmn.tkn.txt" rel="nofollow">中文字符词元化示例</a></li> <li><a href="./desc/addenda/LDC2012T24.eng.raw.txt" rel="nofollow">英文原始译文本示例</a></li> <li><a href="./desc/addenda/LDC2012T24.eng.tkn.txt" rel="nofollow">英文词元化示例</a></li> <li><a href="./desc/addenda/LDC2012T24.wa.txt" rel="nofollow">基于字符的词对齐示例</a></li> </ul><h3>资助说明</h3> <p>本项目部分受美国国防高级研究计划局（Defense Advanced Research Projects Agency, DARPA）GALE项目资助（项目编号：HR0011-06-1-0003）。本文档内容不一定代表美国政府的立场或政策，不应被视为获得官方背书。</p> <h3>更新记录</h3> <p>暂无最新更新。</p> </br> 部分内容 © 2008、2009、2012 宾夕法尼亚大学托管委员会