<!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 class="moz-text-html" lang="x-western">
<div align="center"><b>The Linguistic Data
Consortium (LDC) would like to report on recent developments and
announce the availability of two new publications.</b><br>
<b>
</b><br>
<b>
</b><a
href="http://www.ldc.upenn.edu/Membership/Agreements/member_announcement.shtml#1"><b>LDC
Celebrates its Fifteenth Anniversary!</b></a><br>
<br>
<b><a
href="http://www.ldc.upenn.edu/Catalog/CatalogEntry.jsp?catalogId=LDC2006T13">Free
Google Data (Web 1T 5-gram) Available</a></b><br>
<br>
LDC2007T09<br>
<b><a
href="http://www.ldc.upenn.edu/Catalog/CatalogEntry.jsp?catalogId=LDC2007T09">ISI
Chinese-English Automatically Extracted Parallel Text</a></b><br>
<br>
LDC2007V02<br>
<b><a
href="http://www.ldc.upenn.edu/Catalog/CatalogEntry.jsp?catalogId=LDC2007V02">TRECVID
2003 Keyframes & Transcripts</a></b><br>
</div>
<b><br>
<br>
</b>
<hr size="2" width="100%"><b><br>
</b>
<div align="center"><b>LDC Celebrates its Fifteenth Anniversary!<br>
<br>
</b>
<div align="left">April 15, 2007 marked the start of the LDC's 15th
Anniversary year! We have many milestones to celebrate this year
including the growth of our staff to include over 40 full-time
employees and a online catalog that includes over 350 linguistic
databases. Since 1992, no less than 2,300 organizations from over 80
different nations have licensed LDC data. This data has been made
available through donations, funded projects at LDC or elsewhere,
community initiatives, and, increasingly, LDC initiatives. Over the
past fifteen years, the LDC has grown from an organization that shares
existing language technology resources to one that also is at the
forefront of the creating new data resources, software tools, and
standards. <br>
</div>
<br>
<div align="left">As we celebrate throughout the year, look for new
membership offerings
and announcements. And be sure to join us as we count down to the much
anticipated distribution of our 50,000th publication.<br>
<br>
</div>
<b>Free Google Data Available</b><br>
</div>
<br>
<br>
The LDC is pleased to announce that Google Inc. is providing
financial support for the distribution of its Web 1T 5-gram
(LDC2006T13) corpus to universities. As<br>
a result, LDC will make the corpus available at no charge to 50
non-member universities requesting a copy. Shipping and handling
fees are also being covered by Google. Note that quantities are
limited and the Web 1T 5-gram data is a popular publication. We
appreciate Google's
generosity and its interest in supporting language research. To
obtain a free copy, universities will need to sign and submit a copy of
the <b style=""><a
href="http://www.ldc.upenn.edu/Catalog/nonmem_agree/Web_1T_5gram_V1_User_Agreement.html">User
License Agreement for Web 1T 5-gram Version </a><a
href="http://www.ldc.upenn.edu/Catalog/nonmem_agree/Web_1T_5gram_V1_User_Agreement.html">1</a>
</b>. Please email <a class="moz-txt-link-abbreviated" href="mailto:ldc@ldc.upenn.edu">ldc@ldc.upenn.edu</a> with your contact information.<br>
<br>
<br>
<div align="center"><b>New Publications</b><br>
</div>
<b><br>
<br>
</b>(1) <a
href="http://www.ldc.upenn.edu/Catalog/CatalogEntry.jsp?catalogId=LDC2007T09">ISI
Chinese-English Automatically Extracted Parallel Text</a>
consists of Chinese-English parallel sentences, which were extracted
automatically from two monolingual corpora: Chinese Gigaword Second
Edition (LDC2006T02) and English Gigaword Second Edition (LDC2005T12).
The data was extracted from news articles published by Xinhua News
Agency.
<p>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.</p>
<p>For each sentence pair in the corpus the authors 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. The fact that two documents share several
parallel sentences does not necessarily mean the documents are parallel</p>
<p>In order to make this resource useful for research in Machine
Translation (MT), the authors made efforts to detect potential overlaps
between this data and the standard test and development data sets used
by the MT community. <br>
</p>
<br>
<br>
<div align="center"><b>*</b><br>
</div>
<br>
TREC Video Retrieval Evaluation (TRECVID) is sponsored by the
National Institute of Standards and Technology (NIST) to promote
progress in content-based retrieval from digital video via open,
metrics-based evaluation. The keyframes in <a
href="http://www.ldc.upenn.edu/Catalog/CatalogEntry.jsp?catalogId=LDC2007V02">TRECVID
2003 Keyframes
& Transcripts</a><b> </b>were extracted for use in the NIST
TRECVID 2003 Evaluation. The source data used were English language
broadcast programming collected by LDC in 1998 from ABC ("World News
Tonight") and CNN ("CNN Headline News"). <br>
<br>
TRECVID is a laboratory-style evaluation that attempts to model real
world situations or significant component tasks involved in such
situations. In 2003 there were four main tasks with associated tests: <br>
<br>
<ul>
<li>shot boundary determination </li>
</ul>
<ul>
<li>story segmentation <br>
</li>
</ul>
<ul>
<li>high-level feature extraction </li>
</ul>
<ul>
<li>search (interactive and manual) </li>
</ul>
<br>
Shots are fundamental units of video, useful for higher-level
processing. To create the master list of shots, the video was
segmented. The results of this pass are called subshots. Because the
master shot reference is designed for use in manual assessment, a
second pass over the segmentation was made to create the master shots
of at least 2 seconds in length. These master shots are the ones used
in submitting results for the feature and search tasks in the
evaluation. In the second pass, starting at the beginning of each file,
the subshots were aggregated, if necessary, until the current shot was
at least 2 seconds in duration, at which point the aggregation began
anew with the next subshot. <br>
<br>
The keyframes were selected by going to the middle frame of the shot
boundary, then parsing left and right of that frame to locate the
nearest I-Frame. This then became the keyframe and was extracted.
Keyframes have been provided at both the subshot (NRKF) and master shot
(RKF) levels. <br>
<br>
<hr size="2" width="100%"><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 <a
class="moz-txt-link-freetext" href="http://www.ldc.upenn.edu">http://www.ldc.upenn.edu</a></b></pre>
</div>
</div>
</body>
</html>