SDP 2014 & 2015: Broad Coverage Semantic Dependency Parsing

<h3>Introduction</h3><br>
<p>SDP 2014 &amp; 2015: Broad Coverage Semantic Dependency Parsing consists of data, tools, system results, and publications associated with the 2014 and 2015 tasks on Broad-Coverage Semantic Dependency Parsing (<a href="http://sdp.delph-in.net/">SDP</a>) conducted in conjunction with the International Workshop on Semantic Evaluation (<a href="http://alt.qcri.org/semeval2015/">SemEval</a>) and was developed by the SDP task organizers.</p><br>
<p>SemEval is an ongoing series of evaluations of computational semantic analysis systems intended to explore the nature of meaning in language. It evolved from the <a href="http://www.senseval.org/">Senseval</a> word sense disambiguation series to include semantic analysis tasks outside of word sense disambiguation.</p><br>
<h3>Data</h3><br>
<p>SDP 2014 &amp; 2015: Broad Coverage Semantic Dependency Parsing is based on English, Chinese and Czech data from the following resources: Treebank-2 <a href="../../../LDC95T7">LDC95T17</a>, Proposition Bank I <a href="../../../LDC2004T14">LDC2004T14</a>, NomBank v 1.0 <a href="../../../LDC2008T23">LDC2008T23</a> and CCGBank <a href="../../../LDC2005T13">LDC2005T13 </a> (English); Chinese Treebank (e.g., Chinese Treebank 8.0 <a href="../../../LDC2013T21">LDC2013T21</a>) (Chinese); and Prague Dependency Treebank (e.g., Prague Dependency Treebank 2.0, <a href="../../../LDC2006T01">LDC2006T01</a>) (Czech).</p><br>
<p>The results are presented as graphs in three target representations: MRS-Derived Semantic Dependencies (DM), Enju Predicate&ndash;Argument Structures (PAS), and Prague Semantic Dependencies (PSD). As a fourth, additional target representation CCGbank was converted to semantic dependency graphs (in the subdirectory &lsquo;ccd&rsquo;). These graphs are aligned with the graphs released in connection with SDP 2015 for English.</p><br>
<p>The data is divided into three main directories:</p><br>
<ul><br>
<li>&lsquo;2014/&rsquo; &mdash; the data, tools, and system results from Task 8 at SemEval 2014</li><br>
<li>&lsquo;2015/&rsquo; &mdash; the data, tools, and system results from Task 18 at SemEval 2015</li><br>
<li>&lsquo;ccd/&rsquo; &mdash; the new set of semantic dependency graphs derived from CCGbank</li><br>
</ul><br>
<p>In the 2014 and 2015 sub-directories, the file layout preserves the original conventions used for data distribution to SemEval participants, so as to make it easy to replicate published results. Each sub-directory (including the new &lsquo;ccd/&rsquo;) provides its own file &lsquo;README.txt&rsquo; with additional instructions.</p><br>
<h3>Samples</h3><br>
<p>Please view these samples:</p><br>
<ul><br>
<li><a href="desc/addenda/LDC2016T10.2014.train.txt">2014 Training Data</a></li><br>
<li><a href="desc/addenda/LDC2016T10.2015.graph.pdf">2015 Graph</a></li><br>
<li><a href="desc/addenda/LDC2016T10.ccd.train.txt">CCD Training Data</a></li><br>
</ul><br>
<h3>Updates</h3><br>
<p>None at this time.</p></br>
Portions © 1987-1989 Dow Jones & Company, Inc., © 1996-2005 Institute of Formal and Applied Linguistics and Center for Computational Linguistics, Faculty of Mathematics and Physics, Charles University, © 2005 Julia Hockenmaier and Mark Steedman, © 1995, 1999, 2004, 2005, 2006, 2008, 2013, 2016 Trustees of the University of Pennsylvania

### 简介
<p>SDP 2014与2015年宽覆盖语义依存句法分析（Broad-Coverage Semantic Dependency Parsing，SDP 2014 & 2015）数据集包含与2014、2015年宽覆盖语义依存句法分析任务相关的数据、工具、系统评测结果及学术文献。该任务与国际语义评测研讨会（International Workshop on Semantic Evaluation，SemEval）联合举办，由SDP任务组委会开发。</p>
<p>国际语义评测研讨会（SemEval）是一系列持续进行的计算语义分析系统评测活动，旨在探索语言中的意义本质。该系列评测脱胎于Senseval词语义消歧评测（Senseval word sense disambiguation），现已拓展至词语义消歧以外的各类语义分析任务。</p>
### 数据
<p>SDP 2014与2015年宽覆盖语义依存句法分析数据集基于英语、汉语与捷克语数据构建，数据来源如下：英语数据来自Treebank-2（LDC95T17）、命题库I（Proposition Bank I，LDC2004T14）、NomBank v1.0（LDC2008T23）与CCGBank（LDC2005T13）；汉语数据来自汉语树库（Chinese Treebank，例如汉语树库8.0，LDC2013T21）；捷克语数据来自布拉格依存树库（Prague Dependency Treebank，例如布拉格依存树库2.0，LDC2006T01）。</p>
<p>评测结果以三种目标表征的图结构形式呈现：基于最小递归语义（Minimal Recursion Semantics，MRS）衍生的语义依存关系（MRS-Derived Semantic Dependencies，DM）、Enju谓词-论元结构（Enju Predicate–Argument Structures，PAS）以及布拉格语义依存关系（Prague Semantic Dependencies，PSD）。此外，CCGBank被转换为语义依存图结构，作为第四种目标表征，相关数据存储于子目录‘ccd’中。该子目录中的英语语义依存图与SDP 2015发布的英语语义依存图保持对齐。</p>
<p>数据集分为三个主要子目录：</p>
<ul>
<li>‘2014/’：对应SemEval 2014年第8任务的数据、工具及系统评测结果</li>
<li>‘2015/’：对应SemEval 2015年第18任务的数据、工具及系统评测结果</li>
<li>‘ccd/’：从CCGBank衍生的全新语义依存图数据集</li>
</ul>
<p>2014与2015子目录保留了面向SemEval参赛方的原始数据分发格式，便于研究者复现已发表的评测结果。每个子目录（包括新增的‘ccd/’目录）均附带专属的‘README.txt’文件，包含详细使用说明。</p>
### 示例
<p>请查看以下示例：</p>
<ul>
<li><a href="desc/addenda/LDC2016T10.2014.train.txt">2014训练集</a></li>
<li><a href="desc/addenda/LDC2016T10.2015.graph.pdf">2015年语义依存图</a></li>
<li><a href="desc/addenda/LDC2016T10.ccd.train.txt">CCD训练集</a></li>
</ul>
### 更新
<p>暂无更新。</p>
<p>部分内容版权所有 © 1987-1989 道琼斯公司（Dow Jones & Company, Inc.）、© 1996-2005 查理大学数学与物理学院形式与应用语言学研究所及计算语言学中心、© 2005 Julia Hockenmaier与Mark Steedman、© 1995、1999、2004、2005、2006、2008、2013、2016 宾夕法尼亚大学托管方。</p>

搜集汇总

背景与挑战

背景概述

SDP 2014 & 2015: Broad Coverage Semantic Dependency Parsing 是一个语义依存解析数据集，基于英语、中文和捷克语的数据源（如Treebank-2和Prague Dependency Treebank），包含2014年和2015年SemEval任务的数据、工具和系统结果。该数据集以多种表示形式（如DM、PAS和PSD）呈现语义依存图，适用于解析、语义角色标注和信息提取等自然语言处理应用。