Conferences, Lectures, & Seminars
Events for May
-
NL Seminar-
Fri, May 09, 2014 @ 03:00 PM - 04:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Aram Galstyan, USC/ISI
Talk Title: Deciphering Social Interactions from Text
Series: Natural Language Seminar
Abstract: Studies of social systems have traditionally focused on analyzing various structural properties of networks induced by social communication, while ignoring the content of communication. Despite recent advances, language-based analysis of social processes is still a challenging problem due to the lack of sound mathematical frameworks and adequate computational methods for extracting and analyzing useful social signals from unstructured text. Here I will describe our recent work on content-based analysis of social interactions, which involves two main steps: (a) Embedding communication content in an abstract content space, so that a sequence of textual exchanges is represented as trajectories in this space; and (b) Applying tools from information theory and dynamical systems to discover and characterize directional correlations among those trajectories. I will briefly describe the main elements of the technical approach, and demonstrate the usefulness of the proposed framework on two case studies: content-based characterization of social influence, and stylistic coordination in dialogues.
Biography: Aram Galstyan is a Project Leader at the USC Information Sciences Institute and a Research Assistant Professor at the USC Computer Science Department. His current research focuses on characterizing and predicting behavior of dynamic networks using informationâ⬓theoretic concepts. His other research interests include developing statisticalâ⬓physics based approaches for understanding fundamental limits of various inference algorithms and characterizing the performance of those algorithms with respect to stability and robustness
Host: Aliya Deri and Kevin Knight
More Info: http://nlg.isi.edu/nl-seminar/
Location: Information Science Institute (ISI) - 11th Flr Conf Rm # 1135, Marina Del Rey
Audiences: Everyone Is Invited
Contact: Peter Zamar
Event Link: http://nlg.isi.edu/nl-seminar/
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
NL Seminar- Qualification Practice Talk / Beyond Parallel Data
Wed, May 14, 2014 @ 03:00 PM - 04:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Qing Dou, USC/ISI
Talk Title: Beyond Parallel Data
Series: Natural Language Seminar
Abstract: Thanks to the availability of parallel data and advances in machine learning techniques, we have seen tremendous improvement in the field of machine translation over the past 20 years. However, due to lack of parallel data, the quality of machine translation is still far from satisfying for many language pairs and domains. In general, it is easier to obtain non-parallel data, and much work has tried to learn translations from non-parallel data. Nonetheless, improvements to machine translation have been limited. In this work, I follow a decipherment approach to learn translations from non parallel data and achieve significant gains in machine translation.
I apply slice sampling to Bayesian decipherment. Compared with the state-of-the-art algorithm, the new approach is highly scalable and accurate, making it possible to decipher billions of tokens with hundreds of thousands of word types at high accuracy for the first time. Furthermore, I introduce dependency relations to address the problems of word reordering, insertion, and deletion when deciphering foreign languages, and show that dependency relations help improve deciphering accuracy by over 5-fold. I decipher large amounts of monolingual data to learn translations for out-of-vocabulary words and observe significant gains of up to 3.8 BLEU points in domain-adaptation. Moreover, I show that a translation lexicon learned from large amounts of non-parallel data with decipherment can improve a phrase-based machine translation system trained with limited parallel data. In experiments, I observe BLEU gains of 1.2 to 1.8 across three different test sets.
Given the above success, I propose to work on advancing machine translation of real world low density languages, and to explore using non-parallel data to improve word alignment and discovery of phrase translations.
Qing Dou is a fourth year PhD student at USC/ISI, advised by Professor Kevin Knight.
Biography: Home Page:
http://www.isi.edu/~qdou/
Host: Aliya Deri and Kevin Knight
More Info: http://nlg.isi.edu/nl-seminar/
Location: Information Science Institute (ISI) - 11th Flr Conf Rm # 1135, Marina Del Rey
Audiences: Everyone Is Invited
Contact: Peter Zamar
Event Link: http://nlg.isi.edu/nl-seminar/
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
NL Seminar- Story-Level Inference to Improve Machine Reading
Fri, May 16, 2014 @ 03:00 PM - 04:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Hans Chalupsky, USC/ISI
Talk Title: Story-Level Inference to Improve Machine Reading
Series: Natural Language Seminar
Abstract: Extracting well-defined entities and relations that hold between them from unstructured text is an important prerequisite for a variety of tasks such as knowledge base population, question answering, data analytics, visualization, etc. The difficulty of this problem is evidenced by the annual TAC-KBP evaluations organized by NIST, where the best-performing systems in the slot-filling task still only achieve an f-value in the high 30's. These high error rates on individual relations get further compounded once relations have to be joined to answer a question.
State-of-the art statistical information extraction techniques focus primarily on the phrase and sentence level to extract entities and relations between them, and are generally ignorant of the greater context around them. We present a new approach which aggregates locally extracted information into a larger story context and uses abductive reasoning to generate the best story-level interpretation. We demonstrate that this approach can significantly improve relation extraction and question answering performance on complex questions. We will also describe ongoing work to apply this type of inference to the TAC Knowledge Base Population task in order to improve relation extraction and coreference resolution.
Biography: Hans Chalupsky is a project leader at the Information Sciences Institute of the University of Southern California, where he leads the Loom Knowledge Representation and Reasoning Group. He holds a Master's degree in computer science from the Vienna University of Technology, Austria and a Ph.D. in computer science from the State University of New York at Buffalo. Dr. Chalupsky has over 25 years of experience in the design, development and application of knowledge representation and reasoning systems such as PowerLoom, and he is the principal architect of the KOJAK Link Discovery System. His research interests include knowledge representation and reasoning systems, natural language processing, knowledge and link discovery, anomaly detection and semantic interoperability.
Host: Aliya Deri and Kevin Knight
More Info: http://nlg.isi.edu/nl-seminar/
Location: Information Science Institute (ISI) - 11th Flr Conf Rm # 1135, Marina Del Rey
Audiences: Everyone Is Invited
Contact: Peter Zamar
Event Link: http://nlg.isi.edu/nl-seminar/
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
NL Seminar- How to Speak a Language Without Knowing It
Fri, May 23, 2014 @ 03:00 PM - 04:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Xing Shi, USC/ ISI
Talk Title: How to Speak a Language Without Knowing It
Series: Natural Language Seminar
Abstract: We develop a system that lets people overcome language barriers by letting them speak a language they do not know. Our system accepts text entered by a user, translates the text, then converts the translation into a phonetic spelling in the userââ¬â¢s own orthography. We trained the system on phonetic spellings in travel phrasebooks.
Biography: Xing Shi is a PhD student at USC, advised by Professor Kevin Knight.
Host: Aliya Deri and Kevin Knight
More Info: http://nlg.isi.edu/nl-seminar/
Location: Information Science Institute (ISI) - 11th Flr Conf Rm # 1135, Marina Del Rey
Audiences: Everyone Is Invited
Contact: Peter Zamar
Event Link: http://nlg.isi.edu/nl-seminar/
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
AI Seminar- Frank Schweitzer: Modeling User Behavior In Online Social Networks
Tue, May 27, 2014 @ 03:00 PM - 04:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Frank Schweitzer, Chair of Systems Design, ETH Zurich, Switzerland
Talk Title: Modeling User Behavior In Online Social Networks
Series: Artificial Intelligence Seminar
Abstract: Online communication can be seen as a large-scale social experiment
that constantly provides us with data about users' activities,
interactions and emotions. While their online behavior on the
``microôô level is largely governed by individual traits, we find on the ``macroôô level remarkable statistical regularities. These can be reproduced by means of agent-based models that allow us to understand how emotional influence and individual decisions create collective phenomena, such as collective emotions and drop-out cascades of users.
Biography: Home Page:
http://www.sg.ethz.ch/team/people/fschweitzer/
Host: Kristina Lerman
Webcast: http://webcasterms1.isi.edu/mediasite/Viewer/?peid=6103774e5c2c4eedb6dfc8d4f19ebe471dLocation: Information Science Institute (ISI) - 6th Flr Conf Rm # 689, Marina Del Rey
WebCast Link: http://webcasterms1.isi.edu/mediasite/Viewer/?peid=6103774e5c2c4eedb6dfc8d4f19ebe471d
Audiences: Everyone Is Invited
Contact: Peter Zamar
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
AI SEMINAR
Fri, May 30, 2014 @ 11:00 AM - 12:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Manuel Gomez Rodriguez, Ph.D (MPI)
Talk Title: Modeling Diffusion of Competing Products and Conventions in Social Media
Series: AISeminar
Abstract: Abstract: The emergence and wide-spread use of social networks and
microblogging sites has led to a dramatic increase on the availability
of users' activity data. Importantly, this data can be exploited to
solve some of the problems that have captured the attention of
economists and marketers for decades as, e.g., product adoption,
product competition and product life cycle. In our work, we leverage
on users' activity data from a popular microblogging site to model and
predict the competing dynamics of products and social conventions
adoptions.
To this aim, we propose a data-driven model, based on continuous-time
Hawkes processes, for the adoption of competing products and
conventions. We then develop an inference method to efficiently fit
the model parameters by solving a convex program. The problem
decouples into a collection of smaller subproblems, thus scaling
easily to networks with hundred of thousands of nodes. We validate our
method over synthetic and real diffusion data gathered from Twitter,
and show that the proposed model does not only present a good
predictive power but also provides interpretable model parameters,
which allow us to gain insights into the fundamental principles that
drive product and convention adoptions.
Biography: Bio: Manuel Gomez Rodriguez is a Research Scientist at Max Planck
Institute for Intelligent Systems. Manuel develops machine learning
and large-scale data mining methods for the analysis and modeling of
large real-world networks and processes that take place over them. He
is particularly interested in problems motivated by the Web and social
media and has received several recognitions for his research,
including an Outstanding Paper Award at NIPS'13 and a Best Research
Paper Honorable Mention at KDD'10. Manuel holds a PhD in Electrical
Engineering from Stanford University and a BS in Electrical
Engineering from Carlos III University in Madrid (Spain). You can find
more about him at http://people.tuebingen.mpg.de/manuelgr/
Host: Greg Ver Steeg
Webcast: tbaLocation: Information Science Institute (ISI) - 1135
WebCast Link: tba
Audiences: Everyone Is Invited
Contact: Alma Nava / Information Sciences Institute
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor.
