<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
<div align="center">The Linguistic Data
Consortium (LDC) would like to
announce the availability of
two new publications and provide information regarding
forthcoming publications. <br>
</div>
<p align="center">LDC2007S03<b><br>
<a
href="http://www.ldc.upenn.edu/Catalog/CatalogEntry.jsp?catalogId=LDC2007S03">ARL
Urdu Speech Database, Training Data</a><br>
</b></p>
<div align="center">
</div>
<p align="center">LDC2007T08<br>
<b><a
href="http://www.ldc.upenn.edu/Catalog/CatalogEntry.jsp?catalogId=LDC2007T08">ISI
Arabic-English Automatically Extracted Parallel Text</a><br>
</b></p>
<p align="center"><b>TRECVID Data Update<br>
</b></p>
<p align="center"><b>2007 Publications Pipeline<br>
</b></p>
<hr size="2" width="100%">
<div align="center"><b><br>
New Publications<br>
<br>
</b></div>
<p>(1) <a
href="http://www.ldc.upenn.edu/Catalog/CatalogEntry.jsp?catalogId=LDC2007S03">ARL
Urdu Speech Database, Training Data</a>, is a collection of recorded
speech from 200 adult native Urdu speakers from Pakistan and Northern
India. The database is divided into two parts, a training set
containing approximately 80% of the data and a test set comprised of
20% of the data. This release consists of approximately 80% of the
complete dataset (training and test). The recordings in this release
were collected by Appen Pty Ltd, Sydney, Australia in 2006. </p>
<p>Each speaker was presented with 400 prompts to read: sentences,
place names, and person names. Two microphones set at different
distances to the speaker were used for the recordings. The recorded
speech was stored in raw format files with headers stored in separate
directories. </p>
<p>Each utterance is transcribed in the corresponding label file for
each recording. The transcriptions were encoded in UTF-8. Punctuation
was omitted and numbers were written out in full. <br>
</p>
<br>
<p align="center">*<br>
</p>
<p>(2) <a
href="http://www.ldc.upenn.edu/Catalog/CatalogEntry.jsp?catalogId=LDC2007T08">ISI
Arabic-English Automatically Extracted Parallel Text</a> consists of
Arabic-English parallel sentences which were extracted automatically
from two monolingual corpora: Arabic Gigaword Second Edition
(LDC2006T02) and English Gigaword Second Edition (LDC2005T12). The data
was extracted from news articles published by Xinhua News Agency and
Agence France Presse. The corpus contains 1,124,609 sentence pairs;
the word count on the English side is approximately 31M words. The
sentences in the parallel corpus preserve the form and encoding of the
texts in the original Gigaword corpora.</p>
<p>For each sentence pair in the corpus we provide the names of the
documents from which the two sentences were extracted, as well as a
confidence score (between 0.5 and 1.0), which is indicative of their
degree of parallelism. The parallel sentence identification approach is
designed to judge sentence pairs in isolation from their contexts, and
can therefore find parallel sentences within document pairs which are
not parallel. <br>
</p>
<p>In order to make this resource useful for research in Machine
Translation (MT), we made efforts to detect potential overlaps between
this data and the standard test and development data sets used by the
MT community. <br>
</p>
<p class="MsoNormal" align="center"><br>
</p>
<p class="MsoNormal" align="center"><b>TRECVID Data Update<br>
<br>
</b></p>
<p class="MsoNormal">We've received many queries about the TRECVID data
and are working on a plan to make all of this data available in the LDC
catalog. We anticipate releasing the keyframes for TRECVID 2003 and
2005 later this year. Please watch our website for future
announcements.
<br>
<br>
</p>
<p class="MsoNormal" align="center"><b>2007 Publications Pipeline<br>
<br>
</b></p>
<p class="MsoNormal" align="left"><font
face="Times New Roman, Times, serif">Membership Year (MY) 2007 is
projected to be another productive one for the LDC. In addition to the
aforementioned TRECVID keyframes data, we anticipate releasing a
diverse and exciting selection of publications. Here is a glimpse of
what is in the pipeline for MY 2007</font><font face="Courier New"><small><font
face="Times New Roman, Times, serif">.<big> (Disclaimer: unforeseen
circumstances may lead to modifications of our plans. Please regard
this list as tentative).</big></font>
</small></font></p>
<ul>
<li>GALE Year 1 - Chinese Broadcast Audio, Part 1- first portion of
Mandarin Chinese audio collected for the DARPA GALE Program, including
broadcast news plus talk shows, roundtable discussions and other
conversational news genres. </li>
</ul>
<ul>
<li>GALE Year 1 - Chinese Broadcast Transcripts, Part 1 - manual and
web-harvested transcripts corresponding to the audio included in the
GALE Year 1 Chinese Broadcast Audio corpus. A subset of the
transcripts include both verbatim transcription and manual SU
(sentence-unit) identification plus other rich markup.</li>
</ul>
<ul>
<li>ISI Chinese-English Automatically Extracted Parallel Text<b> -</b>
Chinese-English parallel sentences, which were extracted automatically
from two monolingual corpora: Chinese Gigaword Second Edition
(LDC2006T02) and English Gigaword Second Edition (LDC2005T12). The
corpus contains 558,567 sentence pairs; the word count on the English
side is approximately 16M words. The sentences in the parallel corpus
preserve the form and encoding of the texts in the original Gigaword
corpora. </li>
</ul>
<ul>
<li>OntoNotes V 1.0 - English and Chinese broadcast news transcripts
annotated for Treebank, PropBank, coreference and related information.</li>
</ul>
<ul>
<li>Spoken Levantine Arabic Treebank - experimental pilot annotation
developed for the Johns
Hopkins University
Center for Language and Speech Processing Summer Workshop (WS'05). The
corpus covers morphological and syntactic
annotations of approximately 26,000 words of Levantine Arabic
conversational
telephone speech and was developed under severe time constraints.
Issues of morphological definitions of dialectal words, phrases and
collocations were central to the whole linguistic
description. Syntactic annotation focused on annotation of disfluencies
and on new verbal paradigm and new
structures (e.g., the use of present/active participles).</li>
</ul>
<ul>
<li>Tagged Chinese Gigaword - fully segmented and POS-tagged version
of
Chinese Gigaword Second Edition (LDC2005T14). The CKIP Segmentation and
POS tags were applied uniformly to all texts regardless of its origin.
The size of this tagged corpus, after compression, is about 1.53 GB.</li>
</ul>
As a reminder, MY 2006 will remain open for joining through December
31, 2007 and MY 2007 through December 31, 2008. Organizations may join
for a future MY at any time.<br>
<br>
<hr size="2" width="100%"><br>
<br>
<div align="center"><small><font face="Courier New, Courier, monospace">Ilya
Ahtaridis<br>
Membership Coordinator</font></small><br>
--------------------------------------------------------------------
<font face="Courier New, Courier, monospace"><br>
</font></div>
<div align="center">
<pre class="moz-signature" cols="72"><b><small><font
face="Courier New, Courier, monospace">
</font></small>Linguistic Data Consortium Phone: (215) 573-1275
University of Pennsylvania Fax: (215) 573-2175
3600 Market St., Suite 810 <a
class="moz-txt-link-abbreviated" href="mailto:ldc@ldc.upenn.edu">ldc@ldc.upenn.edu</a>
Philadelphia, PA 19104 USA <a
class="moz-txt-link-freetext" href="http://www.ldc.upenn.edu">http://www.ldc.upenn.edu</a></b></pre>
</div>
</body>
</html>