Conferences, Lectures, & Seminars
Events for February
-
NL Seminar-Deciphering Dark Web through k-partite Graph Summarization
Fri, Feb 05, 2016 @ 03:00 PM - 04:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Linhong Zhu, USC/ISI
Talk Title: Deciphering Dark Web through k-partite Graph Summarization
Series: Natural Language Seminar
Abstract: Facts and their relations extracted from web are commonly modeled as graphs with different types of vertices. In this work, we focus on the problem of revealing latent entities from a $k$-partite graph, by co-clustering $k$ types of different vertices. We propose a CoSum approach, which creates a summary graph, where each super node (a cluster of original vertices) represents a hidden entity and the weighted edges encode important relations among extracted entities. The resulted summary graph also allows for investigation and interpretation of hidden entities. Evaluation verifies that CoSum outperforms several baselines in terms of entity coherence, query supporting and recovering hidden victims in the applied human trafficking domain.
Biography: Linhong Zhu is currently a computer scientist at Information Sciences Institute, University of Southern California, where she also received training as a Postdoctoral Research Associate. Before that, she worked as a Scientist-I in data analytics department at Institute for Infocomm Research, Singapore. She obtained her Ph.D. degree in computer engineering from Nanyang Technological University, Singapore in 2011. Her research interests are large-scale graph analytics with applications to social network analysis, social media analysis, and predictive modeling. She has been awarded with University of Southern California Postdoctoral travel and training award in 2014 and her paper has been selected as two of the best papers in SIGMOD 2010.
Host: Xing Shi and Kevin Knight
More Info: http://nlg.isi.edu/nl-seminar/
Location: Information Science Institute (ISI) - 6th Flr Conf Rm # 689, 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. -
Multi-scale integration and modularity in complex dynamical systems
Fri, Feb 12, 2016 @ 11:00 AM - 12:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Artemy Kolchinsky, Santa Fe Institute
Talk Title: AI Seminar-Multi-scale integration and modularity in complex dynamical systems
Series: Artificial Intelligence Seminar
Abstract: I will discuss two novel approaches to studying distributed organization in complex dynamical systems. In the first [1], we define an information-theoretic measure of the strength of integration at multiple scales, where scale is defined according to an underlying distance metric. We show that our method generalizes several existing complexity measures and is tractable to compute. As demonstrated on human resting state fMRI time-series data, it also captures important aspects of integration in network- and spatially-embedded systems.
In the second approach [2], we address modularity, a pattern of organization in which a system is composed of weakly-coupled subsystems. We develop a technique to decompose dynamical systems based on the idea that modules constrain the spread of perturbations. The method captures variation of modular organization across different system states, time scales, and in response to different kinds of perturbations. It also offers a principled alternative to community detection applied to statistical-dependency networks (e.g. correlation matrices or "functional networks").
[1] A Kolchinsky, MP van den Heuvel, A Griffa, P Hagmann, LM Rocha, O Sporns and J Goñi, Multi-scale Integration and Predictability in Resting State Brain Activity, Frontiers Neuroinformatics, 2014. http://journal.frontiersin.org/article/10.3389/fninf.2014.00066/abstract
[2] A Kolchinsky, AJ Gates, and LM Rocha, Modularity and the spread of perturbations in complex dynamical systems, PRE, 2015. http://arxiv.org/abs/1509.04386
Biography: Artemy Kolchinsky received his PhD from the Center for Complex Systems and Networks, Dept of Informatics, Indiana University Bloomington in 2015. He is currently a postdoctoral fellow at the Santa Fe Institute, collaborating on projects involving optimal compression of dynamical systems as well as thermodynamic constraints on computation. He is broadly interested in novel methods for understanding multivariate dynamics in application to computational neuroscience, evolutionary biology, and other complex systems.
Host: Greg Ver Steeg
Webcast: http://webcasterms1.isi.edu/mediasite/Viewer/?peid=f413ecae075e40eaa3f6b51123178b791dLocation: Information Science Institute (ISI) - 11th Flr Conf Rm # 1135, Marina Del Rey
WebCast Link: http://webcasterms1.isi.edu/mediasite/Viewer/?peid=f413ecae075e40eaa3f6b51123178b791d
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. -
Facebook's Datacenter and Backbone Networks
Fri, Feb 12, 2016 @ 12:00 PM - 01:20 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Dr. Rishi Sinha, Facebook
Talk Title: Facebook's Datacenter and Backbone Networks
Abstract: This talk will cover the design, operational, performance and capacity issues in global networking for large online services, using Facebook as a case study. We will describe Facebook's datacenter and backbone network architecture, explain the characteristics and unique demands of traffic generated in serving a billion daily users, detail the motivations for Facebook's decisions to adopt a next-generation fabric network architecture and to design its own network switches and accompanying operating system, and provide insights into the protocol and software engineering work that is applied to solving performance and capacity challenges in Facebook's network. Finally, we will point to open areas for research and commercialization.
Biography: Dr. Sinha is a performance capacity engineer at Facebook and leads several projects on server capacity planning, network capacity planning, efficiency, and data center logistics. Prior to joining Facebook at 2012, he worked at Brocade where he developed analysis tools for flow control bottlenecks in storage networks, and at Akamai where he worked on reliability of real-time streaming. He has extensive experience in packet flow analysis, experimentation and implementation of internet-scale systems and has four patents on networking related technologies. Dr. Sinha is a Trojan and completed his PhD at USC in 2006.
Host: Alefiya Hussain
Location: Mark Taper Hall Of Humanities (THH) - 210
Audiences: Everyone Is Invited
Contact: Alefiya Hussain
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-Recent Advances in Neural Machine Translation
Fri, Feb 12, 2016 @ 03:00 PM - 04:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Thang Luong, Stanford University
Talk Title: Recent Advances in Neural Machine Translation
Series: Natural Language Seminar
Abstract: Neural Machine Translation (NMT) is a simple new architecture for getting machines to learn to translate. At its core, NMT is a single big recurrent neural network that is trained end-to-end with several advantages such as simplicity and generalization. Despite being relatively new, NMT has already been showing promising results in various translation tasks. In this talk, I will give an overview of NMT and highlight my recent work on (a) how to address the rare word problem in NMT, (b) how to improve the attention (alignment) mechanism, and (c) how to leverage data from other modalities to improve translation.
Biography: Thang Luong is currently a 5th-year PhD student in the Stanford NLP group under Prof. Chris Manning. In the past, he has published papers on various different NLP-related areas such as digital library, machine translation, speech recognition, parsing, psycholinguistics, and word embedding learning. Recently, his main interest shifts towards the area of deep learning using sequence to sequence models to tackle various NLP problems, especially neural machine translation. He has built state-of-the-art (academically) neural machine translation systems both at Google and at Stanford.
Host: Xing Shi and Kevin Knight
More Info: http://nlg.isi.edu/nl-seminar/
Location: Information Science Institute (ISI) - 6th Flr Conf Rm # 689, 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 Chasing vaccination in social media: Narrative discovery from an unstructured corpus of text
Fri, Feb 19, 2016 @ 03:00 PM - 04:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Ehsan Ebrahimzadeh , UCLA
Talk Title: Chasing vaccination in social media: Narrative discovery from an unstructured corpus of text
Series: Natural Language Seminar
Abstract: The measles outbreak in California was a serious public health crisis. Health officials attributed the outbreak to the increasing number of children whose parents had secured exemptions from vaccination for various vaccine preventable diseases VPDs. We believe that exemption seeking is part of a broader culture of distrust driven in large part by stories circulating in social media. An under-standing of the dynamics of this broader culture is necessary if we are to develop health policies that do not simply address outcomes but rather the cultural basis for decisions leading to those outcomes. We reveal the dynamics of exemption seeking and the greater culture of distrust endemic to these sites by developing a generative statistical mechanical model where stories are represented as net- works with actants such as parents, medical professionals, and religious institutions as nodes, and their various relationships as edges. We estimate the latent but unknown stories circulating on these sites by modeling the posts as a sampling of the hidden story graph. Working with a data set of 2 million posts crawled from parenting sites over a 5 year period, we uncover a strong, persistent story signal in which parents, driven by a distrust of government and medical institutions, devise strategies to secure exemptions for their children from required vaccinations. In these stories, it is the vaccines and not the VPDs that pose a threat to the children. Our method of analyzing social media conversations and the exchange of stories at scale can provide an alert mechanism to health officials, help lay the groundwork for devising community-specific messaging interventions, and inform policy making.
Biography: Ehsan Ebrahimzadeh is a PhD candidate in the Electrical Engineering Department of UCLA, where he is simultaneously working towards my his degree in Applied Mathematics. Broadly speaking, he is interested in Statistics, Applied probability, and Data Analytics. Before joining UCLA in 2013, he received his MASc degree in Electrical Engineering from University of Waterloo, and BSc degrees in Mathematics and Electrical Engineering from Isfahan University of Technology.
Host: Xing Shi 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
Fri, Feb 26, 2016 @ 11:00 AM - 12:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Jie Xu, Assistant Professor at University of Miami
Talk Title: Real time knowledge discovery and decision making
Series: AI Seminar
Abstract: As the world becomes ever more connected and instrumented, decision makers have ever more rapid access to ever changing and growing streams of data, but this makes the decision makers problems ever more complex as well, because it is impossible to learn everything in the time frame in which decisions must be made. What the decision maker must do, therefore, is to discover in real time what is relevant in the enormous stream of data and use the relevant information to make good decisions. This talk presents a systematic framework and associated algorithms that enable a decision maker to do this, and shows how to use them in real time traffic prediction as an application scenario. One key challenge in traffic prediction is how much to rely on prediction models that are constructed using historical data in real time traffic situations. Our decision framework learns from the current traffic situation in real time and predicts the future traffic by matching the current situation to the most effective prediction model. The algorithms we propose yield strong performance guarantees for both the long run and the short run. The applications are numerous besides traffic prediction, including patient monitoring, surveillance, social networks etc.
Biography: Jie Xu is an Assistant Professor in the Department of Electrical and Computer Engineering at the University of Miami. His research mainly focuses on game theory and learning theory. His interests lie in both developing the theory in these areas and applying it in real world engineering systems, including communication networks, cyber-security systems, online social platforms and healthcare informatics. Jie received his BS and MS degrees in Electronic Engineering from Tsinghua University in China in 2008 and 2010, respectively, and a PhD degree in Electrical Engineering from University of California, Los Angeles (UCLA) in 2015. In 2014, he was with IBM T. J. Watson Research Center where he interned in the Clinical Stream Analytics team during the summer. Jie is a recipient of Distinguished PhD Dissertation Award at UCLA.
Webcast will be LIVE Broadcast ONLY (no recording):
http://webcasterms1.isi.edu/mediasite/Viewer/?peid=f1e4af9691fe4b2882a98b305b02458c1d
Host: Linhong Zhu
Location: Information Science Institute (ISI) - 11th floor Large CR
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. -
NL Seminar-Interactive scene design using natural language
Fri, Feb 26, 2016 @ 03:00 PM - 04:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Angel Chang , (Stanford University)
Talk Title: Interactive scene design using natural language
Series: Natural Language Seminar
Abstract: Designing 3D scenes is currently a creative task that requires significant expertise and effort in using complex 3D design interfaces. This design process starts in contrast to the easiness with which people can use language to describe real and imaginary environments. We present an interactive text to 3D scene generation system that allows a user to design 3D scenes using natural language. A user provides input text from which we extract explicit constraints on the objects that should appear in the scene. Given these explicit constraints, the system then uses a spatial knowledge base learned from an existing database of 3D scenes and 3D object models to infer an arrangement of the objects forming a natural scene matching the input description. Using textual commands the user can then iteratively refine the created scene by adding, removing, replacing, and manipulating objects.
Biography: Angel Chang recently received her PhD after working in the Stanford NLP group where she was advised by Chris Manning. Her research focuses on the intersection of natural language understanding, computer graphics, and AI. She is currently a visiting expert at Tableau Research. More details at http://stanford.edu/~angelx/
Host: Xing Shi and Kevin Knight
More Info: http://nlg.isi.edu/nl-seminar/
Webcast: http://webcasterms1.isi.edu/mediasite/Viewer/?peid=735bfbb4ba1a4b749fe591958f837ccb1dLocation: Information Science Institute (ISI) - 11th Flr Conf Rm # 1135, Marina Del Rey
WebCast Link: http://webcasterms1.isi.edu/mediasite/Viewer/?peid=735bfbb4ba1a4b749fe591958f837ccb1d
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.
