<div dir="ltr">------------------------------------------------------------------------<br>Arabic-L: Fri 17 May 2013<br>Moderator: Dilworth Parkinson <<a href="mailto:dilworth_parkinson@byu.edu" target="_blank">dilworth_parkinson@byu.edu</a>><br>
[To post messages to the list, send them to <a href="mailto:arabic-l@byu.edu" target="_blank">arabic-l@byu.edu</a>]<br>[To unsubscribe, send message from same address you subscribed from to<br><a href="mailto:listserv@byu.edu" target="_blank">listserv@byu.edu</a> with first line reading:<br>
unsubscribe arabic-l ]<br><br>-------------------------Directory------------------------------------<br><br>1) Subject:LDC GALE & MADCAT Phase 2 Training Set (Arabic Handwriting)<br>
<br>-------------------------Messages-----------------------------------<br>1)<br>Date: 17 May 2013<br>From:<span name="Linguistic Data Consortium" style="font-size:13px;font-family:arial,sans-serif">Linguistic Data Consortium</span><span style="font-family:arial,sans-serif;font-size:13px;white-space:nowrap"> </span><span style="font-family:arial,sans-serif;font-size:13px;white-space:nowrap">ldc@ldc.upenn</span><br>
Subject:LDC GALE & MADCAT Phase 2 Training Set (Arabic Handwriting)<br><br><p style="font-family:arial,sans-serif;font-size:13px"><br><a name="13eb4b98f0bb7002_13ea953743435679_gale"></a>(1) <a href="http://www.ldc.upenn.edu/Catalog/CatalogEntry.jsp?catalogId=LDC2013T10" target="_blank">GALE Arabic-English Parallel Aligned Treebank -- Newswire</a> (LDC2013T10) was developed by LDC and contains 267,520 tokens of word aligned Arabic and English parallel text with treebank annotations. This material was used as training data in the DARPA GALE (Global Autonomous Language Exploitation) program. Parallel aligned treebanks are treebanks annotated with morphological and syntactic structures aligned at the sentence level and the sub-sentence level. Such data sets are useful for natural language processing and related fields, including automatic word alignment system training and evaluation, transfer-rule extraction, word sense disambiguation, translation lexicon extraction and cultural heritage and cross-linguistic studies. With respect to machine translation system development, parallel aligned treebanks may improve system performance with enhanced syntactic parsers, better rules and knowledge about language pairs and reduced word error rate.<u></u><u></u></p>
<p style="font-family:arial,sans-serif;font-size:13px">In this release, the source Arabic data was translated into English. Arabic and English treebank annotations were performed independently. The parallel texts were then word aligned. The material in this corpus corresponds to the Arabic treebanked data appearing in Arabic Treebank: Part 3 v 3.2 (<a href="http://www.ldc.upenn.edu/Catalog/catalogEntry.jsp?catalogId=LDC2010T08" target="_blank">LDC2010T08</a>) (ATB) and to the English treebanked data in English Translation Treebank: An-Nahar Newswire (<a href="http://www.ldc.upenn.edu/Catalog/catalogEntry.jsp?catalogId=LDC2012T02" target="_blank">LDC2012T02</a>).<u></u><u></u></p>
<p style="font-family:arial,sans-serif;font-size:13px">The source data consists of Arabic newswire from the Lebanese publication An Nahar collected by LDC in 2002. All data is encoded as UTF-8. A count of files, words, tokens and segments is below.<u></u><u></u></p>
<table border="1" cellpadding="0" style="font-family:arial,sans-serif;font-size:13px"><tbody><tr><td style="padding:0.75pt"><p>Language<u></u><u></u></p></td><td style="padding:0.75pt"><p>Files<u></u><u></u></p>
</td><td style="padding:0.75pt"><p>Words<u></u><u></u></p></td><td style="padding:0.75pt"><p>Tokens<u></u><u></u></p></td><td style="padding:0.75pt"><p>Segments<u></u><u></u></p></td></tr><tr><td style="padding:0.75pt">
<p>Arabic<u></u><u></u></p></td><td style="padding:0.75pt"><p>364<u></u><u></u></p></td><td style="padding:0.75pt"><p>182,351<u></u><u></u></p></td><td style="padding:0.75pt"><p>267,520<u></u><u></u></p>
</td><td style="padding:0.75pt"><p>7,711<u></u><u></u></p></td></tr></tbody></table><p style="font-family:arial,sans-serif;font-size:13px"><br>Note: Word count is based on the untokenized Arabic source and token count is based on the ATB-tokenized Arabic source.<u></u><u></u></p>
<p style="font-family:arial,sans-serif;font-size:13px">The purpose of the GALE word alignment task was to find correspondences between words, phrases or groups of words in a set of parallel texts. Arabic-English word alignment annotation consisted of the following tasks:<u></u><u></u></p>
<blockquote style="font-family:arial,sans-serif;font-size:13px"><p>Identifying different types of links: translated (correct or incorrect) and not translated (correct or incorrect)<u></u><u></u></p></blockquote><blockquote style="font-family:arial,sans-serif;font-size:13px">
<p>Identifying sentence segments not suitable for annotation, e.g., blank segments, incorrectly-segmented segments, segments with foreign languages<u></u><u></u></p></blockquote><blockquote style="font-family:arial,sans-serif;font-size:13px">
<p>Tagging unmatched words attached to other words or phrases<u></u><u></u></p></blockquote><p style="font-family:arial,sans-serif;font-size:13px">GALE Arabic-English Parallel Aligned Treebank -- Newswire is distributed via web download.<u></u><u></u></p>
<p style="font-family:arial,sans-serif;font-size:13px">2013 Subscription Members will automatically receive two copies of this data on disc. 2013 Standard Members may request a copy as part of their 16 free membership corpora. Non-members may license this data for US$$1750.<br>
<u></u><u></u></p><p align="center" style="font-family:arial,sans-serif;font-size:13px">*<u></u><u></u></p><p style="font-family:arial,sans-serif;font-size:13px"><a name="13eb4b98f0bb7002_13ea953743435679_madcat"></a>(2) <a href="http://www.ldc.upenn.edu/Catalog/CatalogEntry.jsp?catalogId=LDC2013T09" target="_blank">MADCAT Phase 2 Training Set</a> (LDC2013T09) contains all training data created by LDC to support Phase 2 of the DARPA MADCAT (Multilingual Automatic Document Classification Analysis and Translation)Program. The data in this release consists of handwritten Arabic documents, scanned at high resolution and annotated for the physical coordinates of each line and token. Digital transcripts and English translations of each document are also provided, with the various content and annotation layers integrated in a single MADCAT XML output.<u></u><u></u></p>
<p style="font-family:arial,sans-serif;font-size:13px">The goal of the MADCAT program is to automatically convert foreign text images into English transcripts. MADCAT Phase 2 data was collected from Arabic source documents in three genres: newswire, weblog and newsgroup text. Arabic speaking scribes copied documents by hand, following specific instructions on writing style (fast, normal, careful), writing implement (pen, pencil) and paper (lined, unlined). Prior to assignment, source documents were processed to optimize their appearance for the handwriting task, which resulted in some original source documents being broken into multiple pages for handwriting. Each resulting handwritten page was assigned to up to five independent scribes, using different writing conditions.<u></u><u></u></p>
<p style="font-family:arial,sans-serif;font-size:13px">The handwritten, transcribed documents were checked for quality and completeness, then each page was scanned at a high resolution (600 dpi, greyscale) to create a digital version of the handwritten document. The scanned images were then annotated to indicate the physical coordinates of each line and token. Explicit reading order was also labeled, along with any errors produced by the scribes when copying the text. The annotation results in GEDI XML output files (gedi.xml), which include ground truth annotations and source transcripts<u></u><u></u></p>
<p style="font-family:arial,sans-serif;font-size:13px">The final step was to produce a unified data format that takes multiple data streams and generates a single MADCAT XML output file with all required information. The resulting madcat.xml file has these distinct components: (1) a text layer that consists of the source text, tokenization and sentence segmentation, (2) an image layer that consist of bounding boxes, (3) a scribe demographic layer that consists of scribe ID and partition (train/test) and (4) a document metadata layer.<u></u><u></u></p>
<p style="font-family:arial,sans-serif;font-size:13px">This release includes 27,814 annotation files in both GEDI XML and MADCAT XML formats (gedi.xml and madcat.xml) along with their corresponding scanned image files in TIFF format.<u></u><u></u></p>
<p style="font-family:arial,sans-serif;font-size:13px">MADCAT Phase 2 Training Set is distributed on six DVD-ROM.<u></u><u></u></p><p style="font-family:arial,sans-serif;font-size:13px">2013 Subscription Members will automatically receive two copies of this data on disc. 2013 Standard Members may request a copy as part of their 16 free membership corpora. Non-members may license this data for US$2500.</p>
<p style="font-family:arial,sans-serif;font-size:13px"><br></p>--------------------------------------------------------------------------<br>End of Arabic-L: 17 May 2013<br></div>