<div dir="ltr"><p class="MsoNormal" style="margin-bottom:0.0001pt;font-size:12.8px;line-height:normal"><b><span style="font-family:verdana,sans-serif;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><font size="4">Distinguished Lecture Series:  </font></span></b></p><p class="MsoNormal" style="margin-bottom:0.0001pt;font-size:12.8px;line-height:normal"><b><span style="font-family:verdana,sans-serif;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><font size="4">Christopher Manning, Stanford</font></span></b><span style="font-size:18pt;font-family:"times new roman",serif"></span></p><p class="MsoNormal" style="margin-bottom:0.0001pt;font-size:12.8px;line-height:normal;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><span style="font-size:18pt"> </span>​<img src="cid:ii_iua9sc4j11_157c5092169dd420" width="188" height="203" class="gmail-m_-4262302555808638860gmail-CToWUd gmail-m_-4262302555808638860gmail-a6T gmail-CToWUd gmail-a6T" tabindex="0" style="margin-right: 0px;"></p><div style="font-size:12.8px">​​<br></div><div><b><span style="font-family:verdana,sans-serif;color:rgb(48,48,48)">Friday, October 21, 2016 at 11:00 am</span></b><br></div><div style="font-size:12.8px"><b><span style="font-family:verdana,sans-serif;color:rgb(48,48,48)"><br></span></b></div><div style="font-size:12.8px"><p class="MsoNormal" style="margin-bottom:0.0001pt;font-size:12.8px;line-height:normal;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><b><i><span style="font-family:verdana,sans-serif;color:rgb(11,83,148)">Toyota Technological Institute at Chicago</span></i></b><span style="font-family:"times new roman",serif"></span></p><p class="MsoNormal" style="margin-bottom:0.0001pt;font-size:12.8px;line-height:normal;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><b><span style="font-family:verdana,sans-serif;color:rgb(48,48,48)">6045 S. Kenwood Avenue</span></b><span style="font-family:"times new roman",serif"></span></p><p class="MsoNormal" style="margin-bottom:0.0001pt;font-size:12.8px;line-height:normal;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><b><span style="font-family:verdana,sans-serif;color:rgb(48,48,48)">Room #526</span></b><b><span style="color:rgb(48,48,48)">​</span></b><b><span style="font-family:verdana,sans-serif;color:rgb(48,48,48)">/530</span></b><span style="font-family:"times new roman",serif"></span></p><p class="MsoNormal" style="margin-bottom:0.0001pt;font-size:12.8px;line-height:normal;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><b><span style="font-family:verdana,sans-serif;color:rgb(48,48,48)">Christopher Manning, PhD</span></b><span style="font-family:"times new roman",serif"></span></p><p class="MsoNormal" style="margin-bottom:0.0001pt;font-size:12.8px;line-height:14.6pt;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><span style="font-family:verdana,sans-serif;color:black">Professor of Linguistics and Computer Science <br>Natural Language Processing Group<br>Stanford University</span><span style="font-family:verdana,sans-serif"> <span style="font-size:12pt"></span></span></p><p class="MsoNormal" style="margin-bottom:0.0001pt;font-size:12.8px;line-height:14.6pt;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><span style="font-size:12pt;font-family:verdana,sans-serif"> </span></p><p class="MsoNormal" style="margin-bottom:0.0001pt;font-size:12.8px;line-height:14.6pt;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><span style="font-size:12pt;font-family:"times new roman",serif"><a href="http://nlp.stanford.edu/manning" target="_blank"><b><font color="#0000ff">homepage</font></b></a></span></p><p class="MsoNormal" style="margin-bottom:0.0001pt;font-size:12.8px;line-height:14.6pt;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><span style="font-size:12pt;font-family:"times new roman",serif"> </span></p><p class="MsoNormal" style="margin-bottom:0.0001pt;font-size:12.8px;line-height:14.6pt;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><b><u><span style="font-family:verdana,sans-serif">Title</span></u></b><span style="font-family:verdana,sans-serif">: Reading Comprehension and Natural Language Inference</span></p><p class="MsoNormal" style="margin-bottom:0.0001pt;font-size:12.8px;line-height:14.6pt;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><span style="font-family:verdana,sans-serif"> </span></p><p class="MsoNormal" style="font-size:12.8px;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><b><u><span style="font-family:verdana,sans-serif">Abstract</span></u></b><span style="font-family:verdana,sans-serif">: </span><span style="font-family:verdana,sans-serif">Much of computational linguistics and text understanding is either towards one end of the spectrum where there is no representation of compositional linguistic structure (bag-of-words models) or near the other extreme where very complex representations are employed (first order logic, AMR, HPSG, ...). A unifying theme of much of my recent work is to explore models with just a little bit of appropriate linguistic structure. I will focus here on recent case studies in reading comprehension and question answering, exploring the use of both natural logic and deep learning methods for reading comprehension and question answering.</span><span style="font-family:verdana,sans-serif"></span></p><p class="MsoNormal" style="margin-bottom:0.0001pt;font-size:12.8px;line-height:normal;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><span style="font-family:verdana,sans-serif">Enabling a computer to understand a document so that it can answer comprehension questions is a central, yet still unsolved goal of NLP. I’ll first introduce our recent work on the Deepmind QA dataset - a recently released dataset of millions of examples constructed from news articles. On the one hand, we show that (simple) neural network models are surprisingly good at solving this task and achieving state-of-the-art accuracies; on the other hand, we did a careful hand-analysis of a small subset of the problems, and we argue that we are quite close to a performance ceiling on this dataset, and it is still quite far from genuine deep / complex understanding .I will then turn to the use of Natural Logic, a weak proof theory on surface linguistic forms which can nevertheless model many of the common-sense inferences that we wish to make over human language material. I will show how it can support common-sense reasoning and be part of a more linguistically based approach to open information extraction which outperforms previous systems. I show how to augment this approach with a shallow lexical classifier to handle situations where we cannot find any supporting premises. With this augmentation, the system gets very promising results on answering 4th grade science questions, improving over both the classifier in isolation, a strong IR baseline, and prior work. Finally, I will look at how we can incorporate more of the compositional structure of language, which is standardly used in logical approaches to understanding, into a deep learning model. I will emphasize some reason work which shows how that can be done quite efficiently by building the structure like a shift-reduce parser, and how the resulting system can produce stronger results than a sequence model on a natural language inference task.</span></p><p class="MsoNormal" style="margin-bottom:0.0001pt;font-size:12.8px;line-height:normal;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><span style="font-family:verdana,sans-serif"> </span></p><p class="MsoNormal" style="margin-bottom:0.0001pt;font-size:12.8px;line-height:normal;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><span style="font-family:verdana,sans-serif">The talk will include joint work with Gabor Angeli, Danqi Chen, and Sam Bowman.</span></p><p class="MsoNormal" style="margin-bottom:0.0001pt;font-size:12.8px;line-height:normal;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><span style="font-family:verdana,sans-serif"> </span></p><p class="MsoNormal" style="font-size:12.8px"><b><u><span style="font-family:verdana,sans-serif">Bio</span></u></b><span style="font-family:verdana,sans-serif">: </span><span style="font-family:verdana,sans-serif">Christopher Manning is a professor of computer science and linguistics at Stanford University. His Ph.D. is from Stanford in 1995, and he held faculty positions at Carnegie Mellon University and the University of Sydney before returning to Stanford. His research goal is computers that can intelligently process, understand, and generate human language material. Manning concentrates on machine learning approaches to computational linguistic problems, including syntactic parsing, computational semantics and pragmatics, textual inference, machine translation, and using deep learning for NLP. He is an ACM Fellow, a AAAI Fellow, and an ACL Fellow, and has coauthored leading textbooks on statistical natural language processing and information retrieval. He is a member of the Stanford NLP group (@stanfordnlp).</span></p><p class="MsoNormal" style="font-size:12.8px"><span style="font-size:12pt;line-height:18.4px"> <br> <br></span><span style="font-family:verdana,sans-serif">Host: Kevin Gimpel, <a href="http://kgimpel@ttic.edu/" target="_blank">kgimpel@ttic.edu</a></span></p><p class="MsoNormal" style="font-size:12.8px"><br></p><p class="MsoNormal" style="font-size:12.8px"><br></p><p class="MsoNormal" style="font-size:12.8px"><br></p><p class="MsoNormal" style="font-size:12.8px"><br></p><p class="MsoNormal" style="margin-bottom:0.0001pt;font-size:12.8px;line-height:normal;background-image:initial;background-position:initial;background-size:initial;background-repeat:initial;background-origin:initial;background-clip:initial"><span style="font-size:12pt;font-family:verdana,sans-serif"> </span></p></div><div><div class="gmail_signature"><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><font face="arial, helvetica, sans-serif">Mary C. Marre</font><div><font face="arial, helvetica, sans-serif">Administrative Assistant</font></div><div><i><font face="arial, helvetica, sans-serif" color="#3d85c6"><b>Toyota Technological Institute</b></font></i></div><div><i><font face="arial, helvetica, sans-serif" color="#3d85c6">6045 S. Kenwood Avenue</font></i></div><div><i><font face="arial, helvetica, sans-serif" color="#3d85c6">Room 504</font></i></div><div><i><font face="arial, helvetica, sans-serif" color="#3d85c6">Chicago, IL  60637</font></i></div><div><i><font face="arial, helvetica, sans-serif">p:(773) 834-1757</font></i></div><div><i><font face="arial, helvetica, sans-serif">f: (773) 357-6970</font></i></div><div><b><i><a href="mailto:mmarre@ttic.edu" target="_blank"><font face="arial, helvetica, sans-serif">mmarre@ttic.edu</font></a></i></b></div></div></div></div></div></div></div></div></div></div>
</div>