Events: Difference between revisions

Computational Linguistics and Information Processing

Line 17: Line 17:
== Spring 2012 Speakers ==
== Spring 2012 Speakers ==

=== Feb 22: Rebecca Hwa, The Role of Machine Translation in Modeling English as a Second Language (ESL) Writings ===
Much of the English text on the web is written by people whose native
language is not English. Many English as a Second Language (ESL)
writers, even those with a high level of proficiency, make common
grammatical mistakes. In this talk, I will present some of our recent
work on modeling ESL writings. I will first talk about characterizing
ESL students’ writing revision process with syntax-driven machine
translation models. Then, I will focus on modeling similarities in ESL
writers’ preposition selection mistakes using a distance metric
learning algorithm.
BIO: Rebecca Hwa is an Associate Professor in the Department of Computer
Science at the University of Pittsburgh. Before joining Pitt, she was
a postdoc at University of Maryland. She received her PhD in Computer
Science from Harvard University in 2001 and her B.S. in Computer
Science and Engineering from UCLA in 1993. Dr. Hwa's primary research
interests include multilingual processing, machine translation, and
semi-supervised learning methods. She is a recipient of the NSF CAREER
Award. Her work has also been supported by NIH and DARPA. Dr. Hwa is a
past chair of the executive board of the North American Chapter of the
Association for Computational Linguistics.

=== Feb 15: Lidan Wang, Learning to Efficiently Rank (AVW 3258) ===
=== Feb 15: Lidan Wang, Learning to Efficiently Rank (AVW 3258) ===

Revision as of 15:29, 14 February 2012


Titles and abstracts appear after the calendar. Talks are held at 11AM in AV Williams 3258 unless otherwise noted. All are welcome. Typically, external speakers have slots for one-on-one meetings with Maryland researchers. Contact the host if you'd like to have a meeting.

If you would like to get on the list or for other questions about the colloquium series, e-mail Jordan Boyd-Graber.

Google Calendar for CLIP Speakers

{{#widget:Google Calendar | |color=B1440E |title=CLIP Events |view=AGENDA |height=300 }}

Spring 2012 Speakers

Feb 22: Rebecca Hwa, The Role of Machine Translation in Modeling English as a Second Language (ESL) Writings

Much of the English text on the web is written by people whose native language is not English. Many English as a Second Language (ESL) writers, even those with a high level of proficiency, make common grammatical mistakes. In this talk, I will present some of our recent work on modeling ESL writings. I will first talk about characterizing ESL students’ writing revision process with syntax-driven machine translation models. Then, I will focus on modeling similarities in ESL writers’ preposition selection mistakes using a distance metric learning algorithm.

BIO: Rebecca Hwa is an Associate Professor in the Department of Computer Science at the University of Pittsburgh. Before joining Pitt, she was a postdoc at University of Maryland. She received her PhD in Computer Science from Harvard University in 2001 and her B.S. in Computer Science and Engineering from UCLA in 1993. Dr. Hwa's primary research interests include multilingual processing, machine translation, and semi-supervised learning methods. She is a recipient of the NSF CAREER Award. Her work has also been supported by NIH and DARPA. Dr. Hwa is a past chair of the executive board of the North American Chapter of the Association for Computational Linguistics.

Feb 15: Lidan Wang, Learning to Efficiently Rank (AVW 3258)

Technological advances have led to increases in the types and amounts of data, and there is a great need for developing methods to manage and find relevant information from such data to satisfy user's information needs. Learning to rank is an emerging discipline at the intersection of machine learning, data mining, and information retrieval. It develops principled machine learning algorithms to construct ranking (i.e., retrieval) models, for finding and ranking relevant information to user queries over large amounts of data. Although learning to rank approaches are capable of learning highly effective ranking functions, they have mostly ignored the important issue of model efficiency (i.e., model speed). Given that efficiency and effectiveness are competing forces that often counteract each other, models that are optimized for effectiveness alone may not meet the strict efficiency requirements when dealing with real-world large-scale datasets.

My Ph.D. thesis introduces the Learning to Efficiently Rank framework for learning large-scale ranking models that facilitate fast and effective retrieval, by exploiting and optimizing the tradeoffs between model complexity (i.e., speed) and accuracy. At a basic level, this framework learns ranking models whose speed and accuracy can be explicitly controlled. I proposed and designed solutions for three problems within this framework: 1) learning large-scale ranking models according to a desired tradeoff between model speed and accuracy; 2) constructing temporally-constrained models capable of returning results under time budgets; 3) breaking through the speed/accuracy tradeoff barrier by developing a novel cascade ranking model, and learning the cascade model structure and parameters with a novel boosting-based learning algorithm. My research extends the conventional effectiveness-centric approach in model learning and takes an efficiency-minded look at building effective retrieval models. Results show that models learned this way significantly outperform traditional machine-learned models in terms of speed without sacrificing result effectiveness. Moreover, the new models work particularly well when users impose stringent time requirements for ranked retrieval on very large datasets.

Bio: I am a final-year Ph.D. student in the Computer Science Department at the University of Maryland, College Park. My main research interests are in information retrieval, machine learning, and text mining, with an emphasis on efficient and scalable methods for managing, learning, and retrieving information from large-scale data. I have a Master's degree from the Computer Science Department, University of Wisconsin, Madison, and a Bachelor's degree from the Computer Science and Engineering Department, University of Florida. My primary advisor is Prof. Jimmy Lin.

Feb 8: Janyce Wiebe, Subjectivity and Sentiment Analysis: from Words to Discourse (AVW 3258)

There is growing interest in the automatic extraction of opinions, emotions, and sentiments in text (subjectivity analysis) to support natural language processing applications, ranging from mining product reviews and summarization, to automatic question answering and information extraction. In this talk, I will describe work on tasks in subjectivity analysis along a continuum: from subjectivity word sense labeling through discourse-level opinion interpretation and stance recognition.

Bio: Janyce Wiebe is Professor of Computer Science and Director of the Intelligent Systems Program at the University of Pittsburgh. Her research with students and colleagues has been in discourse processing, pragmatics, and word-sense disambiguation. A major concentration of her research is "subjectivity analysis", recognizing and interpretating expressions of opinions and sentiments in text, to support NLP applications such as question answering, information extraction, text categorization, and summarization. Her professional roles have included ACL Program Co-Chair, NAACL Program Chair, NAACL Executive Board member, Computational Linguistics and Language Resources and Evaluation Editorial Board member, AAAI Workshop Co-Chair, ACM Special Interest Group on Artificial Intelligence (SIGART) Vice-Chair, and ACM-SIGART/AAAI Doctoral Consortium Chair.

Fall 2011 Speakers

Sept 7, 14: 5-Minute Madness (AVW 2460)

See what everybody has been working on and get to know who's in the lab!

Sept 21: Youngjoong Ko, Comparison Mining (AVW 2460)

Almost every day, people are faced with a situation that they must decide upon one thing or the other. To make better decisions, they probably attempt to compare entities that they are interested in. These days, many web search engines are helping people look for their interesting entities. It is clear that getting information from a large amount of web data retrieved by the search engines is a much better and easier way than traditional survey methods. However, it is also clear that directly reading each document is not a perfect solution. If people only have access to a small amount of data, they may get a biased point of view. On the other hand, investigating large amounts of data is a time-consuming job. Therefore, a comparison mining system, which can automatically provide a summary of comparisons between two (or more) entities from a large quantity of web documents, would be very useful in many areas such as marketing.

In this talk, I will describe how to build a Korean comparison mining system. Our work is composed of two consecutive tasks: 1) classifying comparative sentences into different types, and 2) mining comparative entities and predicates. We performed various experiments to find relevant features and learning techniques. As a result, we achieved outstanding performance enough for practical use.

Bio: Youngjoong Ko is an associate professor of Computer Engineering at Dong-A University in Korea. He received his PhD in 2003 at Sogang University. His research focuses on text mining (opinion mining, text classification/summarization), Information Retrieval, Dialogue System (speech-act analysis, dialogue modeling). He is currently at CLIP laboratory of UMIACS in University of Maryland as a visiting scholar. Homepage:

Sept 28: No Speaker (Rosh Hashana)

Oct 5: Dave Uthus, Overcoming Information Overload in Navy Chat (AVW 2328)

Abstract: In this talk, I will describe the research we are undertaking at the Naval Research Laboratory which revolves around chat (such as Internet Relay Chat) and the problems it causes in the military domain. Chat has become a primary means for command and control communications in the US Navy. Unfortunately, its popularity has contributed to the classic problem of information overload. For example, Navy watchstanders monitor multiple chat rooms while simultaneously performing their other monitoring duties (e.g., tactical situation screens and radio communications). Some researchers have proposed how automated techniques can help to alleviate these problems, but very little research has addressed this problem.

I will give an overview of the three primary tasks that are the current focus of our research. The first is urgency detection, which involves detecting important chat messages within a dynamic chat stream. The second is summarization, which involves summarizing chat conversations and temporally summarizing sets of chat messages. The third is human-subject studies, which involves simulating a watchstander environment and testing whether our urgency detection and summarization ideas, along with 3D-audio cueing, can aid a watchstander in conducting their duties.

Short Bio: David Uthus is a National Research Council Postdoctoral Fellow hosted at the Naval Research Laboratory, where he is currently undertaking research focusing on analyzing multiparticipant chat. He received his PhD (2010) and MSc (2006) from the University of Auckland in New Zealand and his BSc (2004) from the University of California, Davis. His research interests include microtext analysis, machine learning, metaheuristics, heuristic search, and sport scheduling.

Oct 12: AISTATS Paper Clinic (CLIP Lab)

Oct 13: Nate Chambers, Learning General Events for Specific Event Extraction (AVW 3258)

Abstract: There is a wealth of knowledge about the world encoded in written text. How much of this knowledge and in what form is it accessible by today's unsupervised learning systems? There are two primary views that most systems take on interpreting documents: (1) the document primarily describes specific facts, and (2) the document describes general knowledge about "how the world works" through specific descriptions. These two views are largely separated into two subfields within NLP: Information Extraction, and Knowledge Representations/Induction. Information Extraction is mostly concerned with extracting atomic factoids about the world (e.g., Andrew Luck threw three touchdown passes). Knowledge Induction seeks generalized inferences about the world (e.g., Quarterbacks throw footballs). Although the two operate on similar datasets, most systems focus on only one of the two tasks. This talk will describe my efforts over the past few years to merge the goals of both views, performing unsupervised knowledge induction and information extraction in tandem. I describe a model of event schemas that represents common events and their participants (Knowledge Induction), as well as an algorithm that applies this model to extract specific instances of events from newspaper articles (Information Extraction). I will describe my unique learning approach that relies on coreference resolution to learn event schemas, and then will present the first work that performs template-based IE without labeled datasets or prior knowledge.

If time allows, I will also briefly describe my interests in event ordering and temporal reasoning.

Bio: Nate Chambers is an Assistant Professor in Computer Science at the US Naval Academy. He recently graduated with his Ph.D. in CS from Stanford University. His research interests focus on Natural Language Understanding and Knowledge Acquisition from large amounts of text with minimal human supervision. Before attending Stanford, he worked as a Research Associate at the Florida Institute for Human and Machine Cognition, focusing on human-computer interfaces, dialogue systems, and knowledge representation. He received his M.S. in Computer Science from the University of Rochester in 2003, and has published over 20 peer-reviewed articles.

Oct 17: Michael Collins

There has been a long history in combinatorial optimization of methods that exploit structure in complex problems, using methods such as dual decomposition or Lagrangian relaxation. These methods leverage the observation that complex inference problems can often be decomposed into efficiently solvable sub-problems. Thus far, however, these methods are not widely used in NLP.

In this talk I'll describe recent work on inference algorithms for NLP based on Lagrangian relaxation. In the first part of the talk I'll describe work on non-projective parsing. In the second part of the talk I'll describe an exact decoding algorithm for syntax-based statistical translation. If time permits, I'll also briefly describe algorithms for dynamic programming intersections (e.g., the intersection of a PCFG and an HMM), and for phrase-based translation.

For all of the problems that we consider, the resulting algorithms produce exact solutions, with certificates of optimality, on the vast majority of examples; the algorithms are efficient for problems that are either NP-hard (as is the case for non-projective parsing, or for phrase-based translation), or for problems that are solvable in polynomial time using dynamic programming, but where the traditional exact algorithms are far too expensive to be practical.

While the focus of this talk is on NLP problems, there are close connections to inference methods, in particular belief propagation, for graphical models. Our work was inspired by recent work that has used dual decomposition as an alternative to belief propagation in Markov random fields.

This is joint work with Yin-Wen Chang, Tommi Jaakkola, Terry Koo, Sasha Rush, and David Sontag.

Bio: Michael Collins is the Vikram S. Pandit Professor of Computer Science at Columbia University. His research interests are in natural language processing and machine learning. He completed a PhD in computer science from the University of Pennsylvania in December 1998. From January 1999 to November 2002 he was a researcher at AT&T Labs-Research, and from January 2003 until December 2010 he was an assistant/associate professor at MIT. He joined Columbia University in January 2011. Prof. Collins's research has focused on topics including statistical parsing, structured prediction problems in machine learning, and NLP applications including machine translation, dialog systems, and speech recognition. His awards include a Sloan fellowship, an NSF career award, and best paper awards at several conferences: EMNLP (2002 and 2004), UAI (2004 and 2005), CoNLL 2008, and EMNLP 2010.

Oct 19: Taesun Moon, Pull your head out of your task: broader context in unsupervised models (AVW 2328)

abstract: I discuss unsupervised models and how broader context helps in the resolution of unsupervised or distantly supervised approaches. In the first section, I discuss how document boundaries help in two low-level unsupervised tasks that aren't traditionally resolved in terms of documents: unsupervised morphological segmentation/clustering and unsupervised part-of-speech tagging. For unsupervised morphology, I describe an intuitive model that uses document boundaries to strongly constrain how stems may be clustered and segmented with minimal parameter tuning. For unsupervised part-of-speech tagging, I discuss the crouching Dirichlet, hidden Markov model, an unsupervised POS-tagging model which takes advantage of the difference in the statistical variance of content word and function word POS-tags across documents. Next, I discuss a model of inferring probabilistic word meaning as a distribution over potential paraphrases within context. As opposed to many current approaches in lexical semantics which consider a limited subset of words in a sentence to infer meaning in isolation, this model is able to jointly conduct inference over all words in a sentence. Finally, I describe an approach for connecting language and geography that anchors natural language expressions to specific regions of the Earth. The core of the system is a region-topic model, which is used to learn word distributions for each region discussed in a given corpus. This model performs toponym resolution as a by-product, and additionally enables us to characterize a geographic distribution for corpora, individual texts, or even individual words. The last is joint work with Jason Baldridge, Travis Brown, Katrin Erk, and Mike Speriosu at the University of Texas, Austin.

Bio: Taesun Moon received an MA (2009) and PhD (2011) in linguistics from the University of Texas at Austin (2011) under the supervision of Katrin Erk. He received a BA (2002) in English literature from Seoul National University in South Korea.

Oct 27: Tom Griffiths (3:30 p.m., Bioscience Research Building, 1103)

People are remarkably good at acquiring complex knowledge from limited data, as is required in learning causal relationships, categories, or aspects of language. Successfully solving inductive problems of this kind requires having good "inductive biases" -- constraints that guide inductive inference. Viewed abstractly, understanding human learning requires identifying these inductive biases and exploring their origins. I will argue that probabilistic models of cognition provide a framework that can facilitate this project, giving a transparent characterization of the inductive biases of ideal learners. I will outline how probabilistic models are traditionally used to solve this problem, and then present a new approach that uses a mathematical analysis of the effects of cultural transmission as the basis for an experimental method that magnifies the effects of inductive biases.

Nov 2: Jason Eisner, A Non-Parametric Bayesian Approach to Inflectional Morphology (AVW 2460)

We learn how the words of a language are inflected, given a plain text corpus plus a small supervised set of known paradigms. The approach is principled, simply performing empirical Bayesian inference under a straightforward generative model that explicitly describes the generation of

1. The grammar and subregularities of the language (via many finite-state transducers coordinated in a Markov Random Field). 2. The infinite inventory of types and their inflectional paradigms (via a Dirichlet Process Mixture Model based on the above grammar). 3. The corpus of tokens (by sampling inflected words from the above inventory).

Our inference algorithm cleanly integrates several techniques that handle the different levels of the model: classical dynamic programming operations on the finite-state transducers, loopy belief propagation in the Markov Random Field, and MCMC and MCEM for the non-parametric Dirichlet Process Mixture Model.

We will build up the various components of the model in turn, showing experimental results along the way for several intermediate tasks such as lemmatization, transliteration, and inflection. Finally, we show that modeling paradigms jointly with the Markov Random Field, and learning from unannotated text corpora via the non-parametric model, significantly improves the quality of predicted word inflections.

This is joint work with Markus Dreyer.

Bio: Jason Eisner is Associate Professor of Computer Science at Johns Hopkins University, where he is also affiliated with the Center for Language and Speech Processing, the Cognitive Science Department, and the national Center of Excellence in Human Language Technology. He is particularly interested in designing algorithms that statistically exploit linguistic structure. His 80 or so papers have presented a number of algorithms for parsing and machine translation; algorithms for constructing and training weighted finite-state machines; formalizations, algorithms, theorems and empirical results in computational phonology; and unsupervised or semi-supervised learning methods for domains such as syntax, morphology, and word-sense disambiguation.

Nov 3: EACL / WWW Paper Clinic (11AM, CLIP Lab)

Nov 15: Sergei Nirenburg & Marge McShane, Reference Resolution (10AM, AVW 3258)

Most work on reference resolution in natural language processing has been marked by three features: (1) it has concentrated on textual co-reference resolution, which is the linking of text strings with their coreferential, textual antecedents; (2) only a small subset of reference phenomena have been covered – namely, those that are most easily treated by a “corpus annotation + machine learning” development strategy; and (3) the methods used to treat the selected subset do not hold much promise of being extensible to a broader range of more difficult reference phenomena.

Within the theory of Ontological Semantics, we view reference resolution completely differently. For us, resolving reference means linking references of objects and events in a text to their anchors in the fact repository of the system processing the text – or, to use the terminology of intelligent agents, the memory of the agent processing the text. Furthermore, reference relations extend beyond coreference to meronymy, set-member relations, type-instance relations, so-called “bridging” constructions, etc. The result of reference resolution is the appropriate memory modification of the text processing agent.

In this talk we will briefly introduce OntoSem, our semantically-oriented text processing system and then describe the approach to reference resolution used in OntoSem. We will motivate a semantically oriented approach to reference resolution and show how and why it is currently feasible to develop a new generation of reference resolution engines.

Bio: Dr. Sergei Nirenburg is Professor in the CSEE Department AT UMBC and the Director of its Institute for Language and Information Technologies. Before coming to UMBC, Dr. Nirenburg was Director of the Computing Research Laboratory and Professor of Computer Science at New Mexico State University. He received his Ph.D. in Linguistics from the Hebrew University of Jerusalem, Israel, and his M.Sc. in Computational Linguistics from Kharkov State University, USSR. Dr. Nirenburg has written or edited seven books and has published over 130 articles in various areas of computational linguistics and artificial intelligence. Dr. Nirenburg has directed a number of large-scale research and development projects in the areas of natural language processing, knowledge representation, reasoning, knowledge acquisition and cognitive modeling.

Marge McShane is a Research Associate Professor in the Department of Computer Science and Electrical Engineering of UMBC. She received her Ph.D. from Princeton University, with a specialization in linguistics. She works on theoretical and knowledge-oriented aspects of developing language-enabled intelligent agents. She has led several knowledge acquisition and annotation projects, including the development of a general-purpose workbench for developing computationally-tractable descriptions of lesser-studied languages. A special area of Dr. McShane’s interest is reference resolution, particularly its more difficult aspects, such as ellipsis and referential vagueness. She has published two books and over 60 scientific papers.

Nov 30: Claire Monteleoni (AVW 2328): Clustering Algorithms for Streaming and Online Settings

ABSTRACT: Clustering techniques are widely used to summarize large quantities of data (e.g. aggregating similar news stories), however their outputs can be hard to evaluate. While a domain expert could judge the quality of a clustering, having a human in the loop is often impractical. Probabilistic assumptions have been used to analyze clustering algorithms, for example i.i.d. data, or even data generated by a well-separated mixture of Gaussians. Without any distributional assumptions, one can analyze clustering algorithms by formulating some objective function, and proving that a clustering algorithm either optimizes or approximates it. The k-means clustering objective, for Euclidean data, is simple, intuitive, and widely-cited, however it is NP-hard to optimize, and few algorithms approximate it, even in the batch setting (the algorithm known as "k-means" does not have an approximation guarantee). Dasgupta (2008) posed open problems for approximating it on data streams.

In this talk, I will discuss my ongoing work on designing clustering algorithms for streaming and online settings. First I will present a one-pass, streaming clustering algorithm which approximates the k-means objective on finite data streams. This involves analyzing a variant of the k-means++ algorithm, and extending a divide-and-conquer streaming clustering algorithm from the k-medoid objective. Then I will turn to endless data streams, and introduce a family of algorithms for online clustering with experts. We extend algorithms for online learning with experts, to the unsupervised setting, using intermediate k-means costs, instead of prediction errors, to re-weight experts. When the experts are instantiated as k-means approximate (batch) clustering algorithms run on a sliding window of the data stream, we provide novel online approximation bounds that combine regret bounds extended from supervised online learning, with k-means approximation guarantees. Notably, the resulting bounds are with respect to the optimal k-means cost on the entire data stream seen so far, even though the algorithm is online. I will also present encouraging experimental results.

This talk is based on joint work with Nir Ailon, Ragesh Jaiswal, and Anna Choromanska.

BIO: Claire Monteleoni is an assistant professor of Computer Science at George Washington University, and adjunct research faculty at the Center for Computational Learning Systems at Columbia University, where she was previously research faculty. She did a postdoc in Computer Science and Engineering at the University of California, San Diego, and completed her PhD and Masters in Computer Science, at MIT. Her research focus is on machine learning algorithms and theory for problems including learning from data streams, learning from raw (unlabeled) data, learning from private data, and Climate Informatics: accelerating discovery in Climate Science with machine learning. Her papers have received several awards, and she currently serves on the Senior Program Committee of the International Conference on Machine Learning, and the Editorial Board of the Machine Learning Journal.

Dec. 7: Bill Rand: Authority, Trust and Influence: The Complex Network of Social Media (AVW 2328)

The dramatic feature of social media is that it gives everyone a voice; anyone can speak out and express their opinion to a crowd of followers with little or no cost or effort, which creates a loud and potentially overwhelming marketplace of ideas. Given this egalitarian competition, how do users of social media identify authorities in this crowded space? Who do they trust to provide them with the information and the recommendations that they want? Which tastemakers have the greatest influence on social media users? Using agent-based modeling, machine learning and network analysis we begin to examine and shed light on these questions and develop a deeper understanding of the complex system of social media.

Bio: He received his doctorate in Computer Science from the University of Michigan in 2005 where he worked on the application of evolutionary computation techniques to dynamic environments, and was a regular member of the Center for the Study of Complex Systems, where he built a large-scale agent-based model of suburban sprawl. Before coming to Maryland, he was awarded a postdoctoral research fellowship at Northwestern University in the Northwestern Institute on Complex Systems (NICO), where he worked with the NetLogo development team studying agent-based modeling, evolutionary computation and network science.

Spring 2011 Speakers

May 11, Dave Blei: Scalable Topic Modeling

Probabilistic topic modeling provides a suite of tools for the unsupervised analysis of large collections of documents. Topic modeling algorithms can uncover the underlying themes of a collection and decompose its documents according to those themes. This analysis can be used for corpus exploration, document search, and a variety of prediction problems.

In this talk, I will review the state-of-the-art in probabilistic topic models. I will describe the basic ideas behind latent Dirichlet allocation, and discuss a few of the recent topic modeling algorithms that we have developed in my research group.

I will then describe an online strategy for fitting topic models. This approach lets us analyze massive document collections and document collections arriving in a stream. Specifically, we use variational inference to approximate the posterior of the topic model, and we develop a stochastic optimization algorithm for the corresponding objective function. I will describe online algorithms for finite dimensional topic models and for the Bayeisan nonparametric variant based on the hierarchical Dirichlet process.

Our algorithms can fit models to millions of articles in a matter of hours, and I will present a study of 3.3M articles from Wikipedia. These results show that the online approach finds topic models that are as good or better than those found with traditional inference algorithms.

Bio: David Blei is an assistant professor of Computer Science at Princeton University. He received his PhD in 2004 at U.C. Berkeley and was a postdoctoral fellow at Carnegie Mellon University. His research focuses on probabilistic models, Bayesian nonparametric methods, and approximate posterior inference. He works on a variety of applications, including text, images, music, social networks, and scientific data.

May 4, Sinead Williamson: Nonparametric Bayesian models for dependent data

A priori assumptions about the number of parameters required to model our data are often unrealistic. Bayesian nonparametric models circumvent this problem by assigning prior mass to a countably infinite set of parameters, only a finite (but random) number of which will contribute to a given data set. Over recent years, a number of authors have presented dependent nonparametric models -- distributions over collections of random measures associated with values in some covariate space. While the properties of these random measures are allowed to vary across the covariate space, the marginal distribution at each covariate value is given by a known nonparametric distribution. Such distributions are useful for modelling data that vary with some covariate: in image segmentation, proximal pixels are likely to be assigned to the same segment; in modelling documents, topics are likely to increase and decrease in popularity over time.

Most dependent nonparametric models in the literature have Dirichlet process-distributed marginals. While the Dirichlet process is undeniably the most commonly used discrete nonparametric Bayesian prior, this ignores a wide range of interesting models. In my PhD, I have focused on dependent nonparametric models beyond the Dirichlet process -- in particular, on dependent nonparametric models based on the Indian buffet process, a distribution over binary matrices with an infinite number of columns. In this talk, I will give a general introduction to dependent nonparametric models, and describe some of the work I have done in this area.

Bio: Sinead Williamson is a PhD student working with Zoubin Ghahramani at the University of Cambridge, UK. Her main research interests are dependent nonparametric processes and nonparametric latent variable models. She will be visiting the University of Maryland for six months before starting a post doc at Carnegie Mellon University in the Fall.

April 27, Michele Gelfand

In this presentation, I will describe a perspective on metaphor and negotiation that can help to understand, predict, and manage cultural differences in negotiation. The metaphor approach has its roots in linguistics, cognitive science, and cultural psychology. Metaphors are conceptual systems in which different domains of experience are put into the same category so that knowledge from one domain can be used to make sense of the other. Although they have traditionally been conceived of as linguistic devices, metaphors are a basic mechanism through which humans conceptualize experience (Gibbs, 1990; Lakoff, 1987). In the context of negotiation, metaphors serve a number of critical functions in negotiation. First, they function to create negotiators’ subjective intentional realities (Bruner, 1980; Miller, 1997), guiding both thought and action in negotiation. Specifically, metaphors provide a basis for answering the question, "What kind of situation is this? Is it a battle? A game? A dance? A family gathering? A seduction? A visit to the dentist? I will show how metaphoric mappings provide information about what the task is about and dictate specific entailments or scripts that are derived from their source domains. A "Negotiation as Sports" metaphor, for example, suggests a very different task and scripts as compared to a "Negotiation as Dental Work" or a "Negotiation as Marriage” metaphor. Second, shared metaphors function to organize social action in negotiation (Weick, 1979). Through ongoing communicative exchange, negotiators who develop a shared metaphor for negotiation will come to inhabit the same intentional world, will be more organized and "in-sync" in their interactions (Blount & Janicik, 2003), and will be in a better position to negotiate effectively. I also discuss how metaphors for negotiation are selectively developed, activated, and perpetuated through participation in cultural institutions, helping to explain cross-cultural variation in negotiation dynamics, and problems that arise in intercultural negotiations. In the presentation, I will present a number of recent empirical studies, from the lab and the field, and with samples from a number of countries, which provide support for aspects of the theory. I will conclude with a discussion of the role of metaphor in helping create shared reality in intercultural negotiations.

Bio: Dr. Gelfand is a professor in Maryland's psychology department.

April 22, Eugene Agichtein: Mining Rich User Interaction Data to Improve Web Search

Abstract: Web search engines have advanced greatly over the last decade. In particular, query and click logs have been invaluable to understanding and improving searcher experience. Yet, even the immense logs amassed by the major search engines provide only a narrow glimpse into the searcher behavior and goals. I will present novel techniques for acquiring, analyzing, and exploiting a much richer array of searcher interactions including cursor movements, scrolling, and clicks. As a result, we can more accurately infer searcher intent, enabling dramatic improvements for some search tasks. I will also briefly describe a promising medical application of these techniques.

Biosketch: Eugene Agichtein is an Assistant Professor in the Math & CS department at Emory University, where he leads the Intelligent Information Access Lab. Eugene's research centers on Web search and information retrieval, primarily focusing on modeling users interactions in web search and social media to improve access to information on the web. Increasingly, Eugene is collaborating with medical researchers on applications to medical informatics and clinical diagnosis. This work has been supported by NSF, Microsoft Research, HP Labs, Yahoo! Research, and others. More information about Eugene is available at

April 20, Lillian Lee: Language as Influence(d)

What effect does language have on people, and what effect do people have on language?

You might say in response, "Who are you to discuss these problems?" and you would be right to do so; these are Major Questions that science has been tackling for many years. But as a field, I think natural language processing and computational linguistics have much to contribute to the conversation, and I hope to encourage the community to further address these issues. To this end, I'll describe two efforts I've been involved in.

The first project uncovers previously unexamined contextual biases that people may have when determining which opinions to focus on, using helpfulness votes on reviews as a case study to evaluate competing theories from sociology and social psychology. The second project considers linguistic style matching between conversational participants, using a novel setting to study factors that affect the degree to which people tend to instantly adapt to each others' conversational styles.

Joint work with Cristian Danescu-Niculescu-Mizil, Jon Kleinberg, and Gueorgi Kossinets.


Lillian Lee is a professor of computer science at Cornell University. She is the recipient of the inaugural Best Paper Award at HLT-NAACL 2004 (joint with Regina Barzilay), a citation in "Top Picks: Technology Research Advances of 2004" by Technology Research News (also joint with Regina Barzilay), and an Alfred P. Sloan Research Fellowship, and her group's work has been featured in the New York Times. Homepage

April 13, Leora Morgenstern: Knowledge Representation in the DARPA Machine Reading Program

The DARPA Machine Reading Program (MRP) is focused on developing reading systems that serve as a bridge between the informal information found in natural language texts and the powerful AI systems that use formal knowledge. Central to this effort is the integration of knowledge representation and reasoning techniques into standard information retrieval technology.

In this talk, I discuss the knowledge representation components, including the core ontologies and the domain specific reasoning system, for the MRP reading systems. I focus on the spatiotemporal reasoning that serve as the cornerstone for the central challenge of Phase 3 of the Machine Reading Program: building geographical timelines from news reports.


Leora Morgenstern is currently PI of the DARPA Machine Reading evaluation and knowledge infrastructure team at SAIC. Previous to joining SAIC, she spent most of her career at the IBM T.J. Watson Research Center, where she combined foundational AI research with developing cutting-edge and highly profitable applications for Fortune-500 companies. She is noted in particular for her contributions in applying her research in semantic networks, nonmonotonic inheritance networks, and business rules for applications in knowledge management, customer relationship management, and decision support.

Dr. Morgenstern is the author of over forty scholarly publications and holds three patents, which have won several IBM awards due to their value to industry. She has served on the editorial boards of JAIR, AMAI, and ETAI. She has edited several special issues of journals, the most recent of which was a volume of Artificial Intelligence (January 2011) dedicated to John McCarthy's leadership in field of knowledge representation. Together with John McCarthy and Vladimir Lifschitz, she founded the biannual symposium on Logical Formalizations of Commonsense Reasoning, and has served several times as program co-chair of this symposium. She developed and continues to maintain the Commonsense Problem Page, a website devoted to the pursuit of research in formal commonsense knowledge and reasoning.

April 11, Giacomo Inches: Investigating the statistical properties of user-generated documents

The importance of the Internet as a communication medium is reflected in the large amount of documents being generated every day by users of the different services that take place online. We analyzed the properties of some of the established services over the Internet (Kongregate, Twitter, Myspace and Slashdot) and compared them with consolidated collection of standard information retrieval documents (from the Wall Street Journal, Associated Press and Financial Times, as part of the TREC ad-hoc collection). We investigate features such as document similarity, term burstiness, emoticons and Part-Of-Speech analysis, highlighting their similarities and differences.

Giacomo Inches is a Ph.D. student in the Information Retrieval group within the Informatics Faculty at the University of Lugano (Università della Svizzera italiana, USI), Switzerland. His research is focused on short documents analysis using IR, text mining and machine learning techniques of user generated contents like twitter, chat logs, sms and police report archives. He is currently working on the SNF ChatMiner project ("Mining of conversational content for topic identification and author identification."). In prior scientific work he investigated the field of images classification and worked in the field of database systems (RIA, web engineering). Giacomo received his B.Sc. and M.Sc. from the Politecnico di Milano, Italy and hold a Diplom in Informatik from the University of Erlangen-Nuerember, Germany.

April 6, Rachel Pottinger

When heterogeneous databases are combined, they typically have different schemas, i.e., a description of how the data is stored. For information to be shared between these databases, there must be some way for differences in representation to be resolved. Combining these heterogeneous sources so that they can be queried uniformly is known as semantic integration. There are many aspects to semantic integration, including how to create the underlying system that allows queries to be processed to allowing the user to understand the overpowering amount of data available. In this talk, I describe some of the research that my students and I have been doing to increase data utility through semantic integration, particularly when motivated by real world applications.


Rachel Pottinger is an assistant professor in Computer Science at the University of British Columbia. She received her Ph.D. in computer science from the University of Washington in 2004. Her main research interest is data management, particularly semantic data integration, how to manage metadata (data about data), and how to manage data that is currently not well supported by databases.

March 30, Sujith Ravi: Deciphering Natural Language

Current research in natural language processing (NLP) relies heavily on supervised techniques, which require labeled training data. But such data does not exist for all languages and domains. Using human annotation to create new resources is not a scalable solution, which raises a key research challenge: How can we circumvent the problem of limited labeled resources for NLP applications?

Interestingly, cryptanalysts and archaeologists have tackled similar challenges in the past for solving decipherment problems. Our work draws inspiration from these successes and we present a novel, unified decipherment-based approach for solving natural language problems without labeled (parallel) data. In this talk, we show how NLP problems can be modeled as decipherment tasks. For example, in statistical language translation one can view the foreign-language text as a cipher for English.

Combining techniques from classical cryptography and statistical NLP, we then develop novel decipherment methods to tackle a wide variety of problems ranging from letter substitution decipherment to sequence labeling tasks (such as part-of-speech tagging) to language translation. We also introduce novel unsupervised algorithms that explicitly search for minimized models during decipherment and outperform existing state-of-the-art systems on several NLP tasks.

Along the way, we show experimental results on several tasks and finally, we demonstrate the first successful attempt at automatic language translation without the use of bilingual resources. Unlike conventional approaches, these decipherment methods can be easily extended to multiple domains and languages (especially resource-poor languages), thereby helping to spread the impact and benefits of NLP research.


Sujith Ravi is a Ph.D. candidate in Computer Science at the University of Southern California/Information Sciences Institute, working with Kevin Knight. He received his M.S (2006) degree in Computer Science from USC, and a B.Tech (2004) degree in Computer Science from the National Institute of Technology, Trichy in India. He has also held summer research positions at Google Research and Yahoo Research. His research interests lie in natural language processing, machine learning, computational decipherment and artificial intelligence. His current research focuses on unsupervised and semi-supervised methods with applications in machine translation, transliteration, sequence labeling, large-scale information extraction, syntactic parsing, and information retrieval in discourse. Beyond that, his research experience also includes work on cross-disciplinary areas such as theoretical computer science, computational advertising and computer-aided education. During his graduate student career at USC, he received several awards including an Outstanding Research Assistant Award, an Outstanding Teaching Assistant Award, and an Outstanding Academic Achievement Award.

March 16, Mark Liberman: Problems and opportunities in corpus phonetics

Techniques developed for speech and language technology can now be applied as research tools in an increasing number of areas, some of them perhaps unexpected: sociolinguistics, psycholinguistics, language teaching, clinical diagnosis and treatment, political science -- and even theoretical phonetics and phonology. Some applications are straightforward, and the short-term prospects for work in this field are excellent, but there are many interesting problems for which satisfactory solutions are not yet available. In contrast to traditional speech-technology applications areas, in many of these cases the obvious solutions have not been tried.

Bio (from Wikipedia): Mark has a dual appointment at the University of Pennsylvania, as Trustee Professor of Phonetics in the Department of Linguistics, and as a professor in the Department of Computer and Information Sciences. He is the founder and director of the Linguistic Data Consortium. His main research interests lie in phonetics, prosody, and other aspects of speech communication. Liberman is also the founder of (and frequent contributor to) Language Log, a blog with a broad cast of dozens of professional linguists. The concept of the eggcorn was first proposed in one of his posts there.

March 9, Asad Sayeed: Finding Target-Relevant Sentiment Words

A major indicator of the presence of an opinion and its polarity are the words immediately surrounding a potential opinion "target". But not all the words near the target are likely to be relevant to finding an opinion. Furthermore, prior polarity lexica are only of limited value in finding these words given corpora in specialized domains such as the information technology (IT) business press. There is no ready-made labeled data for this genre and no existing lexica for domain-specific polarity words.

This implementation-level talk describes some work in progress in identifying polarity words in an IT business corpus through crowdsourcing, identifying some of the challenges found in multiple failed attempts. We found that annotating at a fine-grained level with trained individuals is slow, costly, and unreliable given articles that are sometimes quite long. In order to crowdsource the task, however, we had to find ways to ask the question that do not require the user to think too hard about exactly what an opinion is and to reduce the propensity to cheat on a difficult question.

We built an CrowdFlower-based interface that uses a drag-and-drop process to classify words in context. We will demonstrate the interface during the talk and show samples of the results, which we are still in the process of gathering. We will also show some of the implementation-level challenges of adapting the CrowdFlower interface to a non-standard UI paradigm.

If there is time, we will also discuss one of the ways in which we plan to use the data through a CRF-based model of the syntactic relationship between sentiment words and target mentions which we developed in FACTORIE and Scala."

Bio: "Asad Sayeed is a PhD candidate in computer science and member of the University of Maryland CLIP lab. He is working on his dissertation in syntactically fine-grained sentiment analysis."

March 2, Ned Talley: An Unsupervised View of NIH Grants - Latent Categories and Clusters in an Interactive Format

The U.S. National Institutes of Health (NIH) consists of twenty-five Institutes and Centers that award ~80,000 grants each year. The Institutes have distinct missions and research priorities, but there is substantial overlap in the types of research they support, which creates a funding landscape that can be difficult for researchers and research policy professionals to navigate. We have created a publicly accessible database ( in which NIH grants are topic modeled using Latent Dirichlet Allocation, and are clustered using a force-directed algorithm for placing grants as nodes in two dimensional space, where they can be accessed in an online map-like format.

Ned Talley is an NIH Program Director who manages grants on synaptic transmission, synaptic plasticity, and advanced microscopy and imaging. For the past two years he has also been focused on NIH grants informatics, in order to address unmet needs at NIH, and to match these needs with burgeoning technologies in artificial intelligence, information retrieval, and information visualization. He has directed this project through collaborations with investigators from University of Southern California, UC Irvine, Indiana University, and University of Massachusetts.

February 16, Ophir Frieder: Humane Computing

Humane Computing is the design, development, and implementation of computing systems that directly focus on improving the human condition or experience. In that light, three efforts are presented, namely, improving foreign name search technology, spam detection algorithms for peer-to-peer file sharing systems, and novel techniques for urinary tract infection treatment.

The first effort is in support of the Yizkor Books project of the Archives Section of the United States Holocaust Memorial Museum. Yizkor Books are commemorative, firsthand accounts of communities that perished before, during, and after the Holocaust. Users of such volumes include historians, archivists, educators, and survivors. Since Yizkor collections are written in 13 different languages, searching them is difficult. In this effort, novel foreign name search approaches which favorably compare against the state of the art are developed. By segmenting names, fusing individual results, and filtering via a threshold, our approach statistically significantly improves on traditional Soundex and n-gram based search techniques used in the search of such texts. Thus, previously unsuccessful searches are now supported.

In the second effort, spam characteristics in peer-to-peer file sharing systems are determined. Using these characteristics, an approach that does not rely on external information or user feedback is developed. Cost reduction techniques are employed resulting in a statistically significant reduction of spam. Thus, the user search experience is improved.

Finally, a novel “self start”, patient-specific approach for the treatment of recurrent urinary tract infections is presented. Using conventional data mining techniques, an approach that improves patient care, reduces bacterial mutation, and lowers treatment cost is presented. Thus, an approach that provides better, in terms of patient comfort, quicker, in terms of outbreak duration, and more economical care for female patients that suffer from recurrent urinary tract infections is described.

Biography Ophir Frieder is the Robert L. McDevitt, K.S.G., K.C.H.S. and Catherine H. McDevitt L.C.H.S. Chair in Computer Science and Information Processing and is Chair of the Department of Computer Science at Georgetown University. His research interests focus on scalable information retrieval systems spanning search and retrieval and communications issues. He is a Fellow of the AAAS, ACM, and IEEE.

February 9, Naomi Feldman: Using a developing lexicon to constrain phonetic category acquisition

Variability in the acoustic signal makes speech sound category learning a difficult problem. Despite this difficulty, human learners are able to acquire phonetic categories at a young age, between six and twelve months. Learners at this age also show evidence of attending to larger units of speech, particularly in word segmentation tasks. This work investigates how word-level information can help make the phonetic category learning problem easier. A hierarchical Bayesian model is constructed that learns to categorize speech sounds and words simultaneously from a corpus of segmented acoustic tokens. No lexical information is given to the model a priori; it is simply allowed to begin learning a set of word types at the same time that it learns to categorize speech sounds. Simulations compare this model to a purely distributional learner that does not have feedback from a developing lexicon. Results show that whereas a distributional learner mistakenly merges several sets of overlapping categories, an interactive model successfully disambiguates these categories. An artificial language learning experiment with human learners demonstrates that people can make use of the type of word-level cues required for interactive learning. Together, these results suggest that phonetic category learning can be better understood in conjunction with other contemporaneous learning processes and that simultaneous learning of multiple layers of linguistic structure can potentially make the language acquisition problem more tractable.

Bio: Naomi was a graduate student in the Department of Cognitive and Linguistic Sciences at Brown University working with Jim Morgan and Tom Griffiths. She's interested in speech perception and language acquisition, especially the relationship between phonetic category learning, phonological development, and perceptual changes during infancy. In January 2011, she became an assistant professor in the Department of Linguistics at the University of Maryland.

February 2, Ahn Jae-wook: Exploratory user interfaces for personalized information access

Personalized information access systems aim to provide tailored information to users according to their various tasks, interests, or contexts. They have long been relied on the ability of algorithms for estimating user interests and generating personalized information. They observe user behaviors, build mental models of the users, and apply the user model for customizing the information. This process can be done even without any explicit user intervention. However, we can add users into the loop of the personalization process, so that the systems can catch user interests even more precisely and the users can flexibly control the behavior of the systems.

In order to exploit the benefits of the user interfaces for personalized information access, we have investigated various aspects of exploratory information access systems. Exploratory information access systems can combine the strengths of algorithms and user interfaces. Users can learn and investigate their information need beyond the simple lookup search strategy. By adding the idea of the exploration to the personalized information access, we could devise advanced user interfaces for the personalization. Specifically, we have tried to understand how we could let users learn, manipulate, and control the core component of many personalized systems, user models. In this presentation, I am going to introduce several ideas about how to present and control user models using different user interfaces. The example studies include open/editable user model, tab-based user model and query control, reference point-based visualization that incorporates the user model and the query spaces, and named-entity based searching/browsing user interface. The results and the lessons of the user studies are discussed.

Bio: Jae-wook Ahn has recently defended his Ph.D. dissertation at the School of Information Sciences, University of Pittsburgh in September 2010. He has worked with his Ph.D. mentor Dr. Peter Brusilovsky and Dr. Daqing He. He is currently a research associate of the Department of Computer Science and the Human Computer Interaction Lab, working with Dr. Ben Shneiderman.

Fall 2010 Speakers

October 20, Kristy Hollingshead: Search Errors and Model Errors in Pipeline Systems

Pipeline systems, in which data is sequentially processed in stages with the output of one stage providing input to the next, are ubiquitous in the field of natural language processing (NLP) as well as many other research areas. The popularity of the pipeline system architecture may be attributed to the utility of pipelines in improving scalability by reducing search complexity and increasing efficiency of the system. However, pipelines can suffer from the well-known problem of "cascading errors," where errors earlier in the pipeline propagate to later stages in the pipeline. In this talk I will make a distinction between two different type of cascading errors in pipeline systems. The first I will term "search errors," where there exists a higher-scoring candidate (according to the model), but that candidate has been excluded from the search space. The second type of error that I will address might be termed "model errors," where the highest-scoring candidate (according to the model) is not the best candidate (according to some gold standard). Statistical NLP models are imperfect by nature, resulting in model errors. Interestingly, the same pipeline framework that causes search errors can also resolve (or work around) model errors; in this talk I will demonstrate several techniques for detecting and resolving search and model errors, which can result in improved efficiency with no loss in accuracy. I will briefly mention the technique of pipeline iteration, introduced in my ACL'07 paper, and introduce some related results from my dissertation. I will then focus on work done with my PhD advisor Brian Roark on chart cell constraints, as published in our COLING'08 and NAACL'09 papers; this work provably reduces the complexity of a context-free parser to quadratic performance in the worst case (observably linear) with a slight gain in accuracy using the Charniak parser. While much of this talk will be on parsing pipelines, I am currently extending some of this work to MT pipelines and would welcome discussion along those lines.

Kristy Hollingshead earned her PhD in Computer Science and Engineering this year, from the Center for Spoken Language Understanding (CSLU) at the Oregon Health & Science University (OHSU). She received her B.A. in English-Creative Writing from the University of Colorado in 2000 and her M.S. in Computer Science from OHSU in 2004. Her research interests in natural language processing include parsing, machine translation, evaluation metrics, and assistive technologies. She is also interested in general techniques on improving system efficiency, to allow for richer contextual information to be extracted for use in downstream stages of a pipeline system. Kristy was a National Science Foundation Graduate Research Fellow from 2004-2007.

October 27, Stanley Kok: Structure Learning in Markov Logic Networks

Statistical learning handles uncertainty in a robust and principled way. Relational learning (also known as inductive logic programming) models domains involving multiple relations. Recent years have seen a surge of interest in the statistical relational learning (SRL) community in combining the two, driven by the realization that many (if not most) applications require both and by the growing maturity of the two fields.

Markov logic networks (MLNs) is a statistical relational model that has gained traction within the AI community in recent years because of its robustness to noise and its ability to compactly model complex domains. MLNs combine probability and logic by attaching weights to first-order formulas, and viewing these as templates for features of Markov networks. Learning the structure of an MLN consists of learning both formulas and their weights.

To obtain weighted MLN formulas, we could rely on human experts to specify them. However, this approach is error-prone and requires painstaking knowledge engineering. Further, it will not work on domains where there is no human expert. The ideal solution is to automatically learn MLN structure from data. However, this is a challenging task because of its super-exponential search space. In this talk, we present a series of algorithms that efficiently and accurately learn MLN structure.

November 1, Owen Rambow: Relating Language to Cognitive State

In the 80s and 90s of the last century, in subdisciplines such as planning, text generation, and dialog systems, there was considerable interest in modeling the cognitive states of interacting autonomous agents. Theories such as Speech Act Theory (Austin 1962), the belief-desire-intentions model of Bratman (1987), and Rhetorical Structure Theory (Mann and Thompson 1988) together provide a framework in which to link cognitive state with language use. However, in general natural language processing (NLP), little use was made of such theories, presumably because of the difficulty at the time of some underlying tasks (such as syntactic parsing). In this talk, I propose that it is time to again think about the explicit modeling of cognitive state for participants in discourse. In fact, that is the natural way to formulate what NLP is all about. The perspective of cognitive state can provide a context in which many disparate NLP tasks can be classified and related. I will present two NLP projects at Columbia which relate to the modeling of cognitive state:

Discourse participants need to model each other's cognitive states, and language makes this possible by providing special morphological, syntactic, and lexical markers. I present results in automatically determining the degree of belief of a speaker in the propositions in his or her utterance.

Bio: PhD from University of Pennsylvania, 1994, working on German syntax. My office mate was Philip Resnik. I worked at CoGentex, Inc (a small company) and AT&T Labs -- Research until 2002, and since then at Columbia as a Research Scientist. My research interests cover both the nuts-and-bolts of languages, specifically syntax, and how language is used in context.

November 10, Bob Carpenter: Whence Linguistic Data?

The empirical approach to linguistic theory involves collecting data and annotating it according to a coding standard. The ability of multiple annotators to consistently annotate new data reflects the applicability of the theory. In this talk, I'll introduce a generative probabilistic model of the annotation process for categorical data. Given a collection of annotated data, we can infer the true labels of items, the prevalence of some phenomenon (e.g. a given intonation or syntactic alternation), the accuracy and category bias of each annotator, and the codability of the theory as measured by the mean accuracy and bias of annotators and their variability. Hierarchical model extensions allow us to model item labeling difficulty and take into account annotator background and experience. I'll demonstrate the efficacy of the approach using expert and non-expert pools of annotators for simple linguistic labeling tasks such as textual inference, morphological tagging, and named-entity extraction. I'll discuss applications such as monitoring an annotation effort, selecting items with active learning, and generating a probabilistic gold standard for machine learning training and evaluation.

November 15, William Webber: Information retrieval effectiveness: measurably going nowhere?

Information retrieval works by heuristics; correctness cannot be formally proved, but must be empirically assessed. Test collections make this evaluation automated and repeatable. Collection-based evaluation has been standard for half a century. The IR community prides itself on the rigour of the experimental tradition that has been built upon this foundation; it is notoriously difficult to publish in the field without a thorough experimental validation. No attention, however, has been paid to the question of whether methodological rigour in evaluation has to verifiable. In this talk, we present a survey of retrieval results published over the past decade, which fails to find evidence that retrieval effectiveness is in fact improving. Rather, each experiment's impressive leap forward is preceded by a few careful steps back.


William Webber is a Research Associate in the Department of Computer Science and Software Engineering at the University of Melbourne, Australia. He has recently completed his PhD thesis, "Measurement in Information Retrieval Evaluation", under the supervision of Professors Alistair Moffat and Justin Zobel.

December 8: Michael Paul: Summarizing Contrastive Viewpoints in Opinionated Text

Performing multi-document summarization of opinionated text has unique challenges because it is important to recognize that the same information may be presented in different ways from different viewpoints. In this talk, we will present a special kind of contrastive summarization approach intended to highlight this phenomenon and to help users digest conflicting opinions. To do this, we introduce a new graph-based algorithm, Comparative LexRank, to score sentences in a summary based on a combination of both representativeness of the collection and comparability between opposing viewpoints. We then address the issue of how to automatically discover and extract viewpoints from unlabeled text, and we experiment with a novel two-dimensional topic model for the task of unsupervised clustering of documents by viewpoint. Finally, we discuss how these two stages can be combined to both automatically extract and summarize viewpoints in an interesting way. Results are presented on two political opinion data sets.

This project was joint work with ChengXiang Zhai and Roxana Girju.

Bio: Michael Paul is a first-year Ph.D. student of Computer Science at the Johns Hopkins University and a member of the Center for Language and Speech Processing. He earned a B.S. from the University of Illinois at Urbana-Champaign in 2009. He is currently a Graduate Research Fellow of the National Science Foundation and a Dean's Fellow of the Whiting School of Engineering.

Roger Levy

Considering the adversity of the conditions under which linguistic communication takes place in everyday life -- ambiguity of the signal, environmental competition for our attention, speaker error, our limited memory, and so forth -- it is perhaps remarkable that we are as successful at it as we are. Perhaps the leading explanation of this success is that (a) the linguistic signal is redundant, (b) diverse information sources are generally available that can help us obtain infer something close to the intended message when comprehending an utterance, and (c) we use these diverse information sources very quickly and to the fullest extent possible. This explanation suggests a theory of language comprehension as a rational, evidential process. In this talk, I describe recent research on how we can use the tools of computational linguistics to formalize and implement such a theory, and to apply it to a variety of problems in human sentence comprehension, including classic cases of garden-path disambiguation as well as processing difficulty in the absence of structural ambiguity. In addition, I address a number of phenomena that remain clear puzzles for the rational approach, due to an apparent failure to use information available in a sentence appropriately in global or incremental inferences about the correct interpretation of a sentence. I argue that the apparent puzzle posed by these phenomena for models of rational sentence comprehension may derive from the failure of existing models to appropriately account for the environmental and cognitive constraints -- in this case, the inherent uncertainty of perceptual input, and humans' ability to compensate for it -- under which comprehension takes place. I present a new probabilistic model of language comprehension under uncertain input and show that this model leads to solutions to the above puzzles. I also present behavioral data in support of novel predictions made by the model. More generally, I suggest that appropriately accounting for environmental and cognitive constraints in probabilistic models can lead to a more nuanced and ultimately more satisfactory picture of key aspects of human cognition.

Earl Wagner

Eugene Charniak

We present a new syntactic parser that works left-to-right and top down, thus maintaining a fully-connected parse tree for a few alternative parse hypotheses. All of the commonly used statistical parsers use context-free dynamic programming algorithms and as such work bottom up on the entire sentence. Thus they only find a complete fully connected parse at the very end. In contrast, both subjective and experimental evidence show that people understand a sentence word-to-word as they go along, or close to it. The constraint that the parser keeps one or more fully connected syntactic trees is intended to operationalize this cognitive fact. Our parser achieves a new best result for top-down parsers of 89.4%,a 20% error reduction over the previous single-parser best result for parsers of this type of 86.8% (Roark01). The improved performance is due to embracing the very large feature set available in exchange for giving up dynamic programming.

Eugene Charniak is University Professor of Computer Science and Cognitive Science at Brown University and past chair of the Department of Computer Science. He received his A.B. degree in Physics from University of Chicago, and a Ph.D. from M.I.T. in Computer Science. He has published four books the most recent being Statistical Language Learning. He is a Fellow of the American Association of Artificial Intelligence and was previously a Councilor of the organization. His research has always been in the area of language understanding or technologies which relate to it. Over the last 20 years years he has been interested in statistical techniques for many areas of language processing including parsing and discourse,

Dave Newman

Ray Mooney

Current systems that learn to process natural language require laboriously constructed human-annotated training data. Ideally, a computer would be able to acquire language like a child by being exposed to linguistic input in the context of a relevant but ambiguous perceptual environment. As a step in this direction, we present a system that learns to sportscast simulated robot soccer games by example. The training data consists of textual human commentaries on Robocup simulation games. A set of possible alternative meanings for each comment is automatically constructed from game event traces. Our previously developed systems for learning to parse and generate natural language (KRISP and WASP) were augmented to learn from this data and then commentate novel games. Using this approach, the system has learned to sportscast in both English and Korean. The system has been evaluated based on its ability to properly match sentences to the events being described, parse sentences into correct meanings, and generate accurate linguistic descriptions of events. Human evaluation was also conducted on the overall quality of the generated sportscasts and compared to human-generated commentaries, demonstrating that its sportscasts are on par with those generated by humans.

Biographical Sketch:

Raymond J. Mooney is a Professor in the Department of Computer Sciences at the University of Texas at Austin. He received his Ph.D. in 1988 from the University of Illinois at Urbana/Champaign. He is an author of over 150 published research papers, primarily in the areas of machine learning and natural language processing. He is the current President of the International Machine Learning Society, was program co-chair for the 2006 AAAI Conference on Artificial Intelligence, general chair of the 2005 Human Language Technology Conference and Conference on Empirical Methods in Natural Language Processing, and co-chair of the 1990 International Conference on Machine Learning. He is a Fellow of the American Association for Artificial Intelligence and recipient of best paper awards from the National Conference on Artificial Intelligence, the SIGKDD International Conference on Knowledge Discovery and Data Mining, the International Conference on Machine Learning, and the Annual Meeting of the Association for Computational Linguistics. His recent research has focused on learning for natural-language processing, connecting language and perception, statistical relational learning, and transfer learning.