-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathspmrl2013-sharedtask.html
More file actions
512 lines (499 loc) · 22 KB
/
spmrl2013-sharedtask.html
File metadata and controls
512 lines (499 loc) · 22 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
<!DOCTYPE html>
<html lang="en">
<head>
<title>Shared Task</title>
<meta charset="utf-8" />
<link rel="stylesheet" href="/theme/css/main.css" type="text/css" />
<link href="/" type="application/atom+xml" rel="alternate" title="Statistical Parsing of Morphologically Rich Languages ATOM Feed" />
<!--[if IE]>
<script src="http://html5shiv.googlecode.com/svn/trunk/html5.js"></script><![endif]-->
<!--[if lte IE 7]>
<link rel="stylesheet" type="text/css" media="all" href="/css/ie.css"/>
<script src="/js/IE8.js" type="text/javascript"></script><![endif]-->
<!--[if lt IE 7]>
<link rel="stylesheet" type="text/css" media="all" href="/css/ie6.css"/><![endif]-->
</head>
<body id="index" class="home">
<header id="banner" class="body">
<h1><a href="/index.html">Statistical Parsing of Morphologically Rich Languages </a>
</h1>
<nav><ul>
<li><a href="/">Archives</a></li>
</ul></nav>
</header><!-- /#banner -->
<section id="content" class="body">
<article>
<header> <h1 class="entry-title"><a href="spmrl2013-sharedtask.html"
rel="bookmark" title="Permalink to Shared Task">Shared Task</a></h1> </header>
<div class="entry-content">
<footer class="post-info">
<abbr class="published" title="2013-10-10T00:00:00">
Do 10 Oktober 2013
</abbr>
<address class="vcard author">
By <a class="url fn" href="/author/the-spmrl-team.html">The SPMRL Team</a>
</address>
<p>In <a href="/category/cfp.html">CfP</a>. </p>
<p>tags: <a href="/tag/cfp.html">CfP</a><a href="/tag/spmrl2013.html">spmrl2013</a></p></p></footer><!-- /.post-info --><!-- /.post-info -->
<!-- -*- coding:utf-8 -*- -->
<p>The fourth SPMRL workshop hosted the first shared task on parsing morphologically rich languages.
The previous page is available at <a class="reference external" href="http://www.spmrl.org/shared_task_old.html">http://www.spmrl.org/shared_task_old.html</a>
(note: this webpage is in heavy editing, will be up to date before the workshop)</p>
<div class="section" id="goals">
<h2>Goals</h2>
<p>The primary goal of the shared task on parsing morphologically rich languages was to bring forward work on parsing morphologically ambiguous input in both dependency and constituency parsing, and to show the state of the art for MRLs. In the longer term, we aim to provide streamlined data sets and evaluation metrics, thus improving the comparability of cross linguistic work on parsing MRLs. The shared task featured
tracks in constituency parsing and in dependency parsing, in gold as well as in realistic scenarios (the realistic scenario has no gold tokenization, no gold part-of-speech tags and morphological features).</p>
</div>
<div class="section" id="data-set">
<h2>Data Set</h2>
<p>The participants were provided with data from 9 different languages (Arabic, Basque, French, German, Hebrew, Hungarian, Korean, Polish,Swedish). The data were available in Penn Treebank bracketing format, CoNLL-X format and optionally in TiGerXML.
In order to ease cross-linguistic comparisons, the data set have also been released within a common size setting (ie, treebanks of 5000 sentences).
All treebanks (dep. and const.) are aligned at the sentence, token and POS levels.</p>
<p>The data set and the procedure to download them are available at</p>
<p><a class="reference external" href="http://dokufarm.phil.hhu.de/spmrl2013/doku.php">http://dokufarm.phil.hhu.de/spmrl2013/doku.php</a></p>
</div>
<div class="section" id="metrics">
<h2>Metrics</h2>
<div class="section" id="gold-tokens-scenarios">
<h3>Gold Tokens Scenarios:</h3>
<ul>
<li><dl class="first docutils">
<dt>Dependency Structure:</dt>
<dd><ul class="first last simple">
<li>We mainly used Labeled Attachment Scores (Buchholz and Marsi, 2006) via CoNLL 2007's <a class="reference external" href="http://pauillac.inria.fr/~seddah/eval07.pl">eval07.pl (with slightly modified output)</a></li>
</ul>
</dd>
</dl>
</li>
<li><p class="first">Phrase Structures</p>
<blockquote>
<ul>
<li><p class="first">We used two metrics: Parseval (Evalb, (Black et al, 91) and LeafAncestor (Sampson and Babarczy, 2003). With a modified version (from Sancl 2012 (Petrov and Mc Donald, 2012) that penalises unparsed trees for the former and with an implementation from Wagner (2012) for the latter.</p>
<blockquote>
<ul class="simple">
<li>Evalb spmrl : <a class="reference external" href="http://pauillac.inria.fr/~seddah/evalb_spmrl2013.tar.gz">evalb_spmrl2013.tar.gz</a> (please read the README file, we used the -X option)</li>
<li>LeafAncestor: <a class="reference external" href="http://pauillac.inria.fr/~seddah/parse_la.py">parse_la.py</a> (please read the disclaimer on top of the file)</li>
</ul>
</blockquote>
</li>
<li><p class="first">Note: as oppposed to the common usage in the parsing communities, all constituency results are given for sentences of all lenght and all tokens are evaluated (including punctuation tokens). For both Evalb and LeafAncestor, the labels {TOP, S1, ROOT, VROOT} are stripped off.</p>
</li>
</ul>
</blockquote>
</li>
<li><p class="first">Multi Word Expressions evaluation:</p>
<blockquote>
<ul>
<li><p class="first">The French data set contains MWEs annotated at the morpho syntactic level. We evaluated them for the dependency track only. (see <a class="reference external" href="http://dokufarm.phil.hhu.de/spmrl2013/doku.php?id=shared_task_description">wiki page</a> )</p>
<blockquote>
<ul class="simple">
<li>MWE Evaluator (written by Marie Candito): <a class="reference external" href="http://pauillac.inria.fr/~seddah/do_eval_dep_mwe.pl">do_eval_dep_mwe.pl</a></li>
</ul>
</blockquote>
</li>
</ul>
</blockquote>
</li>
</ul>
</div>
<div class="section" id="predicted-tokens-scenarios">
<h3>Predicted Tokens Scenarios:</h3>
<ul>
<li><p class="first">Dependency and Constituent Structures</p>
<blockquote>
<ul class="simple">
<li>We used TedEval (Tsarfaty et al 2010,2011,2012) in its realistic framework (namely a test file with its own mapping between predicted tokens and source tokens is evaluated upon a gold file and the gold token mapping). TedEval is available here: <a class="reference external" href="http://www.tsarfaty.com/unipar/download.html">Tedeval 2.2</a>.</li>
<li>We developped a set of wrappers that use MaltParser's reprojectiver (Nivre & Nilsson, 2005). Wrappers are available here: <a class="reference external" href="http://pauillac.inria.fr/~seddah/TedWrappers_20131015.tar.gz">TedWrappers_20131015.tar.gz</a></li>
</ul>
</blockquote>
</li>
</ul>
</div>
</div>
<div class="section" id="results">
<h2>Results</h2>
<p>Background information and extensive details on the evaluation scenarios and discussions on the results are provided on the Shared Task Overview paper (Seddah et al., 2013)
<a class="reference external" href="http://aclweb.org/anthology//W/W13/W13-4917.pdf">http://aclweb.org/anthology//W/W13/W13-4917.pdf</a></p>
<div class="section" id="dependency-parsing-track">
<h3>Dependency Parsing Track</h3>
<p>We used the same protocol as in Conll 2007 (Nivre et al, 2007) in two settings for 4 scenarios:</p>
<p>Full train set size ⇒ with gold or predicted morphology (POS tag and features)
5k sentences train set size ⇒ with gold or predicted morphology (POS tag and features).
Note that the predicted data were provided as baseline, participants were free to use theirs. The French, Hebrew and Arabic predicted train sets have not been subjected to a cross fold jackniffing so participants were incited to do it (only a few did use their own predicted morphology though: Alpage-IGM and Alpage-Dyalog for French, Cadim for Arabic and IMS_SGZEDED_CIS for all languages.)</p>
<ul class="simple">
<li><a class="reference external" href="http://pauillac.inria.fr/~seddah/updated_official.spmrl_results.html">web</a></li>
<li><a class="reference external" href="http://pauillac.inria.fr/~seddah/updated_official.csv">csv</a></li>
</ul>
<div class="section" id="all-languages-ranking-pred-full">
<h4>All languages ranking (Pred/Full)</h4>
<ul class="simple">
<li>1st IMS-SZEGED-CIS</li>
<li>2nd Alpage-Dyalog</li>
<li>3rd MaltOptimier</li>
</ul>
<p>Breakdown by language (Pred/Full)</p>
<ul>
<li><dl class="first docutils">
<dt>Arabic</dt>
<dd><ol class="first last arabic simple">
<li>IMS-SZEGED-CIS</li>
<li>CADIM</li>
<li>MALTOPTIMIZER</li>
</ol>
</dd>
</dl>
</li>
<li><dl class="first docutils">
<dt>Basque</dt>
<dd><ol class="first last arabic simple">
<li>IMS-SZEGED-CIS</li>
<li>BASQUE TEAM</li>
<li>MALTOPTIMIZER</li>
</ol>
</dd>
</dl>
</li>
<li><dl class="first docutils">
<dt>French</dt>
<dd><ol class="first last arabic simple">
<li>IGM-ALPAGE</li>
<li>IMS-SZEGED-CIS</li>
<li>BASQUE TEAM</li>
</ol>
</dd>
</dl>
</li>
<li><dl class="first docutils">
<dt>German</dt>
<dd><ol class="first last arabic simple">
<li>IMS-SZEGED-CIS</li>
<li>BASQUE TEAM</li>
<li>ALPAGE DYALOG</li>
</ol>
</dd>
</dl>
</li>
<li><dl class="first docutils">
<dt>Hebrew</dt>
<dd><ol class="first last arabic simple">
<li>IMS-SZEGED-CIS</li>
<li>ALPAGE DYALOG</li>
<li>MALTOPTIMIZER</li>
</ol>
</dd>
</dl>
</li>
<li><dl class="first docutils">
<dt>Hungarian</dt>
<dd><ol class="first last arabic simple">
<li>IMS-SZEGED-CIS</li>
<li>BASQUE TEAM</li>
<li>MALTOPTIMIZER</li>
</ol>
</dd>
</dl>
</li>
<li><dl class="first docutils">
<dt>Korean</dt>
<dd><ol class="first last arabic simple">
<li>IMS-SZEGED-CIS</li>
<li>MALTOPTIMIZER</li>
<li>BASELINE MALT</li>
</ol>
</dd>
</dl>
</li>
<li><dl class="first docutils">
<dt>Polish</dt>
<dd><ol class="first last arabic simple">
<li>IMS-SZEGED-CIS</li>
<li>ALPAGE DYALOG</li>
<li>MALTOPTIMIZER</li>
</ol>
</dd>
</dl>
</li>
<li><dl class="first docutils">
<dt>Swedish</dt>
<dd><ol class="first last arabic simple">
<li>IMS-SZEGED-CIS</li>
<li>BASQUE TEAM</li>
<li>ALPAGE DYALOG</li>
</ol>
</dd>
</dl>
</li>
</ul>
<p>(Teams that used undisclosed unlabeled data are not part of the official ranking)</p>
</div>
</div>
<div class="section" id="constituent-parsing-track">
<h3>Constituent Parsing track</h3>
<p>Only IMS-SZEGED-CIS participated, we provided baseline results using the Berkeley Parser in its default mode -- s/m: 5, OOV threshold: 10, default lexicon -- (Petrov et al, 2006, 2007) in two modes: with provided POS Tags (ie Tagged) and in Raw mode (the parser did its own pos tagging).</p>
<ul>
<li><dl class="first docutils">
<dt>Parseval scores:</dt>
<dd><ul class="first last simple">
<li><a class="reference external" href="http://pauillac.inria.fr/~seddah/official_ptb-all.spmrl_results.html">parseval web</a></li>
<li><a class="reference external" href="http://pauillac.inria.fr/~seddah/official_ptb-all.csv">parseval csv</a></li>
</ul>
</dd>
</dl>
</li>
<li><dl class="first docutils">
<dt>LeafAncestor scores:</dt>
<dd><ul class="first last simple">
<li><a class="reference external" href="http://pauillac.inria.fr/~seddah/official_leaf-all.spmrl_results.html">leaf web</a></li>
<li><a class="reference external" href="http://pauillac.inria.fr/~seddah/official_leaf-all.csv">leaf csv</a></li>
</ul>
</dd>
</dl>
</li>
</ul>
</div>
<div class="section" id="multi-word-expression-evaluation">
<h3>Multi Word Expression Evaluation</h3>
<ul class="simple">
<li><a class="reference external" href="http://pauillac.inria.fr/~seddah/official-mwe.spmrl_results.html">mwe web</a></li>
<li><a class="reference external" href="http://pauillac.inria.fr/~seddah/mwe_french_eval_conll.csv">mwe csv</a></li>
</ul>
<p>MWE French ranking:</p>
<ul class="simple">
<li>IGM-ALPAGE</li>
<li>IMS-SZEGED-CIS</li>
<li>ALPAGE DYALOG</li>
</ul>
</div>
<div class="section" id="non-gold-token-evaluation">
<h3>Non Gold Token Evaluation</h3>
<p>Arabic and Hebrew data set were provided with generated lattices (disambiguated and non-disambiguated for Hebrew, disambiguated only for Arabic – the data exist though, they should be made available at some points –)</p>
<p>Results on the predicted tokens scenarios are evaluated using Tedeval 2.2 (Tsarfaty et al, 2011,2012) in two modes:</p>
<p>A fully labeled mode (where edges, either from const. trees or dependencies, are decorated by their original labels). This mode allow for a full comparison between dependency parses produced on gold tokens and predicted tokens from the raw source text.
An unlabeled mode which allows for easier cross-framework comparison (between const. and dep. parsers). In order to perform a fully labeled evaluation of a const. tree, each edges needs to bear a function label. Please see the overview paper for full details on the cross framework scenarios.</p>
<ul>
<li><p class="first">Unlabeled mode ( <a class="reference external" href="http://pauillac.inria.fr/~seddah/official_tedeval-unlabel-70.spmrl_results.html">tedeval unlabeled (<70)</a> )</p>
<blockquote>
<ul>
<li><dl class="first docutils">
<dt>Arabic Ranking (full/pred)</dt>
<dd><ol class="first last arabic simple">
<li>IMS-SZEGED-CIS</li>
<li>CADIM</li>
<li>MALTOPTIMER</li>
</ol>
</dd>
</dl>
</li>
<li><dl class="first docutils">
<dt>Hebrew Ranking (full/pred)</dt>
<dd><ol class="first last arabic simple">
<li>IMS-SZEGED-CIS</li>
<li>ALPAGE DYALOG</li>
<li>MALTOPTIMIZER</li>
</ol>
</dd>
</dl>
</li>
</ul>
</blockquote>
</li>
<li><p class="first">Labeled mode ( <a class="reference external" href="http://pauillac.inria.fr/~seddah/official_tedeval-unlabel-70.spmrl_results.html">tedeval labeled (<70)</a> )</p>
<blockquote>
<ul>
<li><dl class="first docutils">
<dt>Arabic Ranking (full/pred)</dt>
<dd><ol class="first last arabic simple">
<li>IMS-SZEGED-CIS</li>
<li>CADIM</li>
<li>MALTOPTIMER</li>
</ol>
</dd>
</dl>
</li>
<li><dl class="first docutils">
<dt>Hebrew Ranking (full/pred)</dt>
<dd><ol class="first last arabic simple">
<li>IMS-SZEGED-CIS</li>
<li>MALTOPTIMIZER</li>
<li>ALPAGE DYALOG</li>
</ol>
</dd>
</dl>
</li>
</ul>
</blockquote>
</li>
</ul>
</div>
</div>
<div class="section" id="getting-the-shared-task-data-set">
<h2>Getting the Shared Task Data Set</h2>
<p>All data but Arabic are freely available under the same conditions as during the shared task.
Unless stated otherwise by their original licenses, any commercial exploitation of treebank data,
derived parsing or tagging models are prohibited. Those data set are made available for
reproductibility's sake and in the hope that this shared task data will provide inspiration
for the design and evaluation of future parsing systems for these languages.</p>
<p>The Arabic data we provided is based on the LDC's ATB 4.1, 3.1 and 3.2, then converted to
both Columbia's CaTib Dependency Treebank (Habash & Roth, 2009) and to Stanford's preprocessed version
of the ATB (Green & Manning, 2010).
It is to be made available soon by the LDC via its usual channels. Contact us at <a class="reference external" href="mailto:spmrl.sharedtask@gmail.com">spmrl.sharedtask@gmail.com</a>
if you absolutely need the data urgently, we'll made available our (huge) set of scripts we developed
to create the data.</p>
<ul class="simple">
<li><a class="reference external" href="http://dokufarm.phil.hhu.de/spmrl2013/doku.php?id=how_to_obtain_licenses_for_the_shared_task_data">Licences</a>.</li>
<li><a class="reference external" href="http://dokufarm.phil.hhu.de/spmrl2013/doku.php?id=frequently_asked_questions#how_are_the_data_organized">How are the Data Organized</a>.</li>
</ul>
</div>
<div class="section" id="shared-task-organizers">
<h2>Shared task Organizers</h2>
<ul class="simple">
<li>Sandra Kübler (Indiana University, US)</li>
<li>Djamé Seddah (Université Paris Sorbonne & INRIAs Alpage Project, France)</li>
<li>Reut Tsarfaty (Weizmann Institute of Science, Israel)</li>
</ul>
</div>
<div class="section" id="contact">
<h2>Contact</h2>
<ul class="simple">
<li>mail: <a class="reference external" href="mailto:spmrl.sharedtask@gmail.com">spmrl.sharedtask@gmail.com</a></li>
<li>Mailing list (still active, do not hesitate to ask questions): <a class="reference external" href="https://sympa.inria.fr/sympa/arc/mrlp-sharedtask">https://sympa.inria.fr/sympa/arc/mrlp-sharedtask</a></li>
</ul>
</div>
<div class="section" id="treebank-provider-teams">
<h2>Treebank Provider Teams</h2>
<ul class="simple">
<li>Coordination effort, Normalization, Alignment between Const. and Dep.:<ul>
<li>Djamé Seddah, Reut Tsarfaty and Sandra Kübler</li>
</ul>
</li>
</ul>
<ul class="simple">
<li>Arabic<ul>
<li>Nizar Habash, Ryan Roth (Columbia University)
<em>Dependencies from Catib's treebank and full predicted morphology data</em></li>
<li>Spence Green (Stanford University)
<em>Code to build Stanford-like Constituents trees</em></li>
<li>Ann Bies, Seth Kullick, Mohammed Maamouri (the Linguistic Data Consortium)
<em>Original data set</em></li>
</ul>
</li>
<li>Basque<ul>
<li>Koldo Gojenola, Iakes Goenaga (University of the Basque Country)</li>
</ul>
</li>
<li>French<ul>
<li>Marie Candito (Univ. Paris 7 & Alpage), Djamé Seddah (Univ. Paris Sorbonne & Alpage)</li>
</ul>
</li>
<li>German<ul>
<li>Wolfgang Seeker (IMS), Wolfgang Maier (Univ. of Dusseldorf), Yannick Versley (Univ. of Tuebingen) & Ines Rehbein (Postdam Univ.)</li>
</ul>
</li>
<li>Hebrew<ul>
<li>Yoav Goldberg (Ben Gurion Univ), Reut Tsarfaty (Weizmann Institute of Science)</li>
</ul>
</li>
<li>Hungarian<ul>
<li>Richárd Farkas, Veronika Vincze (Univ. of Szeged)</li>
</ul>
</li>
<li>Korean<ul>
<li>Jinho D. Choi (IPsoft Inc.)</li>
</ul>
</li>
<li>Polish<ul>
<li>Adam Przepiorkowski, Marcin Woliński, Alina Wróblewska (Institute of Computer Science, Polish Academy of Sciences)</li>
</ul>
</li>
<li>Swedish<ul>
<li>Joakim Nivre (Uppsala Univ.), Marco Kuhlmann (Linköping University)</li>
</ul>
</li>
</ul>
<p>We thank the Linguistic Data Consortium for agreeing to release their current versions of the Arabic Treebank for this Shared Task.
(Ann Bies, Denise Di Pierso, Seth Kullick, Mohammed Maamouri)</p>
<p>Special thanks to Ozlem Cetinoglu & Thomas Müller (IMS), Yuval Marton (Microsoft Inc.), Miguel Ballesteros (Univ. Pompeu Fabra)
for helping us debugging the data set at many occasions!</p>
</div>
<div class="section" id="acknowledgements">
<h2>Acknowledgements</h2>
<p>For their precious help preparing the SPMRL 2013 Shared Task and for
allowing their data to be part of it, we warmly thank the Linguistic
Data Consortium, the Knowledge Center for Processing Hebrew (MILA),
the Ben Gurion University, Columbia University, Institute of Computer
Science (Polish Academy of Sciences), Korea Advanced Institute of
Science and Technology, University of the Basque Country, University
of Lisbon, Uppsala University, University of Stuttgart, University of
Szeged and University Paris Diderot (Paris 7).
We are also very grateful to the Philosophical Faculty of the Heinrich-Heine
Universität Düsseldorf for hosting the shared task data via their dokuwiki.</p>
<p>We take advantage of this page to warmly and publicly thank once more all
the people involved in this shared task preparation (original data
set, scripting, website, institutionnal and moral support):</p>
<p>(no particular order) Marie Candito, Jennifer Foster, Yoav Goldberg,
Ines Rehbein, Yannick Versley, Ann Bies, Denise Di Pierso, Seth
Kullick, Mohammed Maamouri, Spence Green, Christopher Manning, Mona
Diab, Nizar Habash, Yuval Marton, Owen Rambow, Ryan Roth, Jinho
Choi, Key-Sun Choi, Koldo Gojenola, Iakes Goenaga, Richard Farkas,
Veronika Vincze, Adam Przepiorkowski, Alina Wróblewska, Marcin
Woliński, Anne Abeillé, Joseph van Genabith, Marco Kullman, Joakim
Nivre, Ozlem Cetinoglu, Wolfgang Maier, Wolfgang Seeker, Khahil
Sima'an, Harry Bunt, Alon Lavie, Benoit Sagot, Benoit Crabbé,
Laurence Danlos, Eric de la Clergerie, James Henderson, Slav Petrov,
Zhongqiang Huang, Joseph Le Roux, Grzegorz Chrupala.</p>
</div>
</div><!-- /.entry-content -->
</article>
</section>
<aside id="sidebar">
<div class="widget">
<h2>Categories</h2>
<ul>
<li class="active"><a href="/category/cfp.html">CfP</a></li>
<li ><a href="/category/misc.html">misc</a></li>
<li ><a href="/category/program.html">program</a></li>
</ul>
</div>
<div class="widget blogroll">
<h2>External links</h2>
<ul>
<li><a href="http://www.cs.cmu.edu/~sigparse/">SIGParse</a></li>
<li><a href="http://www.aclweb.org/anthology-new/J/J13/#1000">CL special issue</a></li>
<li><a href="https://sites.google.com/site/spsemmrl2012/">SP-Sem-MRL 2012</a></li>
<li><a href="https://sites.google.com/site/spmrl2011/">SPMRL 2011</a></li>
<li><a href="https://sites.google.com/site/spmrl2010/">SPMRL 2010</a></li>
<li><a href="http://alpage.inria.fr/iwpt09/panel.en.html">SPMRL panel (IWPT'09)</a></li>
</ul>
</div><!-- /.blogroll -->
<div class="widget">
<h2>Proceedings</h2>
<ul>
<li>
<a href="http://www.aclweb.org/anthology-new/W/W10/#1400">
SPMRL 2010<br>
<img src="http://pauillac.inria.fr/~seddah/Spmrl.jpg" alt="spmrl 2010 cover" width="200" height="300"></a>
</li>
<li>
<a href="http://www.aclweb.org/anthology-new/W/W11/#3800">
SPMRL 2011<br>
<img src="http://pauillac.inria.fr/~seddah/spmrl2011.jpg" alt="spmrl 2011 cover" width="200" height="300"></a>
</li>
<li>
<a href="http://aclweb.org/anthology-new/W/W12/#3400">SP-Sem-MRL 2012<br/>
<img src="http://pauillac.inria.fr/~seddah/FrontSPSEM2012.jpg" width="201" height="173"/>
</a>
</li>
<li>
<a href="http://aclweb.org/anthology-new/W/W13/#4900">SPMRL 2013<br/>
<img src="http://pauillac.inria.fr/~seddah/FrontSPSEM2012.jpg" width="201" height="173"/>
</a>
</li>
</ul>
</div>
</aside><!-- /#sidebar -->
<footer id="footer" class="body">
<p>Created using <a href="http://alexis.notmyidea.org/pelican/">pelican</a> and
a variation on the «notmyidea-cms» theme, a modified version of «notmyidea», the default theme.</p>
</footer><!-- /#footer -->
</body>
</html>