<html><head></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space; "><br>THEORY, TYPOLOGY & TECHNOLOGY: <br>PARSING IN THE FACE OF DIVERSITY<br>WORKSHOP, MARCH 23, 2010, AMSTERDAM<br><a href="http://www.illc.uva.nl/NewsandEvents/index.php?cat=events#item3272">http://www.illc.uva.nl/NewsandEvents/index.php?cat=events#item3272</a><br><br><div>!!New!! Workshop program has been updated.</div><div><br></div><div>Workshop Abstract:</div><div><br>Any statistical model that automatically analyzes naturally-occurring natural language data must rely on a particular linguistic representation. Different representations and their associated theories embody different assumptions about the linguistic phenomena. Representations may be constituency-based or dependency-based, and they may employ theoretical constructs borrowed from various syntactic formal frameworks (such as GPSG, HPSG, LFG, CCG, TAG and so on). Either way, the representations are assumed to be rich enough to express the empirical facts about the data, so that the model can statistically learn linguistic phenomena, and generalize to new instances.<br><br>When adapting statistical parsing models from one language to another, these assumptions often break down. A representation type that is sufficient for expressing empirical facts about one language is not always rich enough to express facts about a different one, or, on the other hand, it may seem unnecessarily cumbersome. The quest for a general representation formalism that is adequate for a wide range of languages is similar, in spirit, to the search for universal principles that underlie the organization of natural language grammar. Despite the differences in methodology, insights obtained in one of these strands of research may shed new light on the findings of the other.<br><br>This workshop therefore brings together researchers working in language technology, formal grammars, linguistic typology, and theoretical morphology, in order to discuss, compare, contrast and evaluate representation formalisms that are used for modeling natural language phenomena. The idea that morphology competes with syntax is seen as a key to understanding cross-linguistic variation and nonconfigurationality; particular attention will be given to the representation of morphological information and its place in the overall syntactic representation. A desirable outcome of this meeting would also be a better understanding of how data-driven computational approaches can combine with theoretical linguistic analyses to advance our understanding of the principles underlying grammars, and contribute to finding universal tendencies.<br><br>Confirmed Speakers:<br>- Mark Johnson (Macquarie University, Sydney, Australia) <br>- Joakim Nivre (Uppsala University, Sweden) <br>- James P. Blevins (University of Cambridge, UK) <br>- Owen Rambow (Columbia University, USA) <br>- Gregory Stump (University of Kentucky, USA) <br>- Yoad Winter (Utrecht University, The Netherlands) <br>- Reut Tsarfaty (University of Amsterdam, The Netherlands)<br><br>The Computational Linguistics Seminar (CLS) of the Institute for Logic, Language and Computation (University of Amsterdam) is proud to host a one-day workshop on ?Parsing in the face of diversity,? discussing the intersection of theoretical, typological, and technological concerns, on the occasion of the public defence of Reut Tsarfaty?s PhD dissertation, entitled ?Relational-Realizational Parsing?. <br>The workshop will be held on March 23 at the University of Amsterdam, and is organized by Reut Tsarfaty, Stefan Frank, Khalil Sima?an and Remko Scha. Attendance is free and open for all. For planning purposes, please register by March 20. You can register using the following link <a href="http://spreadsheets.google.com/viewform?formkey=dHBYSVd4TnRRT3ZjZ3pfMGZUNnhiaXc6MA">http://spreadsheets.google.com/viewform?formkey=dHBYSVd4TnRRT3ZjZ3pfMGZUNnhiaXc6MA</a><br><br>Workshop Program:<br><br>Tuesday March 23, 2010 </div><div><br>Introduction 9:00?9:45<span class="Apple-tab-span" style="white-space: pre; "> </span><br>Reut Tsarfaty</div><div><br>Session 1: Parsing/Grammars (chair: Remko Scha) <br></div><div><br></div><div>10:00-11:00<span class="Apple-tab-span" style="white-space: pre; "> </span>Joakim Nivre:</div><div>Dependency parsing in the face of diversity</div><div><div>11:00-12:00<span class="Apple-tab-span" style="white-space: pre; "> </span>Mark Johnson:</div><div><div>Bayesian models of language acquisition,</div><div>or, Where do the rules come from?</div><div><br></div></div>Session 2: Morphology/Typology (chair: Reut Tsarfaty)</div><div> <br>13:30-14:30<span class="Apple-tab-span" style="white-space: pre; "> </span>Gregory Stump<br>Metaconjugations and other evidence for a dual representation of inflectional paradigms</div><div>14:30-15:30<span class="Apple-tab-span" style="white-space: pre; "> </span>James P. Blevins:</div><div>Paradigmatic deduction</div><div><br>Session 3: Semitic Syntax/Morphology (chair: Khalil Sima?an) </div><div><br>16:00-17:00<span class="Apple-tab-span" style="white-space: pre; "> </span>Yoad Winter</div><div>Automatic annotation of Morpho-Syntactic dependencies in a Modern Hebrew treebank<br>17:00-18:00<span class="Apple-tab-span" style="white-space: pre; "> </span>Owen Rambow</div><div>Parsing Arabic: More problems than solutions</div><div><br>Concluding discussion 18:15 (everyone)<br><br></div><div>Workshop dinner 19:00<br><br>Workshop Venue:<br><br>Oost Indisch Huis AT.03 <br>Oude Hoogstraat 24 <br>1012 CE Amsterdam<br><br>Questions?<br><br>Email Reut Tsarfaty <a href="mailto:r.tsarfaty@uva.nl">r.tsarfaty@uva.nl</a><br></div></body></html>