<div dir="ltr"><div class="gmail_quote"><div dir="ltr"><div class="gmail_quote"><div><div><div dir="ltr"><div class="gmail_quote"><div dir="ltr"><font color="#000000">CALL FOR PARTICIPATION</font><div>
<font color="#000000"><br></font></div><div><font color="#000000">NLP Unshared Task in PoliInformatics 2014<br><br>Website: <a href="https://sites.google.com/site/unsharedtask2014/" target="_blank">https://sites.google.com/site/unsharedtask2014/</a><br>
<br>Contact: <a href="mailto:unsharedtask@gmail.com" target="_blank">unsharedtask@gmail.com</a><font><div style="font-family:Arial,Verdana,sans-serif;font-size:13px">
<font size="3"><br></font></div></font></font><div><font color="#000000">The financial crisis of 2007-8 was an extremely complex, world-wide event. The U.S. government's response to the crisis was arguably as complex, but better documented. We invite researchers with natural language processing expertise to consider a corpus of reports, hearings, bills, and other transcripts related to the crisis. We have organized a research competition around the data and these questions:<br>
<br>* Who was the financial crisis? We seek to understand the participants in the lawmaking and regulatory processes that formed the government’s response to the crisis: the individuals, industries, and professionals targeted by those policies; the agencies and organizations responsible for implementing them; and the lobbyists, witnesses, advocates, and politicians who were actively involved -- and the connections among them.<br>
<br></font></div><div><font color="#000000">* What was the financial crisis? We seek to understand the cause(s) of the crisis, proposals for reform, advocates for those proposals, arguments for and against, policies ultimately adopted by the government, and the impact of those policies.</font></div>
<font color="#000000"><br>Contrasting with “shared tasks” -- common exercises in the NLP community -- an unshared task does not specify a quantitative performance measure for comparing solutions and does not even specify what a solution might look like. Instead, the organizers provide data and an open-ended prompt. Participants are invited to explore the use of NLP methods to help scholars in political science, communications, and other related fields make sense of a large, complicated corpus. Participants are invited to show what they can do in the form of short papers describing exploratory research and optional demos. We believe many such papers will discuss quantitative and qualitative analysis of existing NLP tools and systems on portions of the data, though new implementations are also welcome, as are newly processed datasets that may be more directly usable in future research projects.</font><div>
<font color="#000000"><br></font><div><font color="#000000">Papers will be reviewed by a panel of judges. These judges will author public responses discussing the relevance of unshared task submissions, suggesting uses in that may be unfamiliar to NLP researchers, as well as new research directions. Above all, an emphasis is placed on evaluating the potential for future interdisciplinary research stemming from unshared task entries. In addition, the panel of judges may present an award to the entry (or entries) with the greatest potential. Our hope is that new collaborations between NLP researchers and those with substantive interests in political science will develop as a result of the unshared task.<br>
<br>Timeline<br><br>* January 1, 2014: official data release<br>* February 14, 2014: deadline to register your team<br>* April 15, 2014: deadline for unshared task submissions <br>* May 31, 2014: deadline for public reviews from the panel of judges<br>
* June 26, 2014: announcement of awardees at the ACL Workshop on Language Technologies and Computational Social Science in Baltimore, Maryland (<a href="http://www.mpi-sws.org/~cristian/LACSS_2014.html" target="_blank">http://www.mpi-sws.org/~cristian/LACSS_2014.html</a>)</font></div>
<div><font color="#000000"><br></font></div><div><font color="#000000">The data are described and can be downloaded at this URL, which also provides a link to register: <a href="https://sites.google.com/site/unsharedtask2014/" target="_blank">https://sites.google.com/site/unsharedtask2014/</a></font></div>
<div><font color="#000000"><span style="font-family:Arial,Verdana,sans-serif;font-size:13px"><br></span></font></div><font color="#000000">Who is organizing this competition?<br><br>PoliInformatics leverages advances in computer science, machine learning, and data visualization to promote analyses of very large and unstructured datasets related to the study of government and politics. The PoliInformatics Research Coordination Network (PInet; <a href="http://poliinformatics.org" target="_blank">http://poliinformatics.org</a>) is a working group funded by the National Science Foundation to build community and capacity for data-intensive research using open government data. PInet has focused its work on the 2007-8 financial crisis, government policy relating to the crisis, and public response to that policy. PInet has provided the data and the panel of judges who will respond to the unshared task entries.</font></div>
<div><font color="#000000"><br></font></div><div><font color="#000000">The NLP unshared task in PoliInformatics is being organized by:<br><br>* Claire Cardie, Cornell University<br>* Noah Smith, Carnegie Mellon University<br>
* Anne Washington, George Mason University<br>
* John Wilkerson, University of Washington
</font></div></div></div>
</div><br></div>
</div></div></div><br></div>
</div><br></div>