Select a calendar:
Filter October Events by Event Type:
Conferences, Lectures, & Seminars
Events for October
-
AI SEMINAR
Fri, Oct 09, 2015 @ 11:00 AM - 12:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Pedro Szekely, Research Associate Professor
Talk Title: Domain Specific Search Where No Search Has Gone Before
Series: AI Seminar
Abstract: We are investigating crawling, extraction, alignment, entity resolution, database and indexing technologies to enable rapid creation of large domain-specific knowledge graphs. We are also investigating analytics, query and visualization techniques that use these graphs to deliver sophisticated, yet easy to use query and analysis capabilities to end-users. Our goal is to build technology that can use any source of information on the Web, including Web pages and services, text, images, text-delimited files and databases, and that scales to 1 billion Web pages. The project is a collaboration of the ISI information integration and natural language processing groups, Columbia University (deep learning for image analysis), JPL (Web crawling), Inferlink (extraction from Web pages and entity resolution) and NextCentury (user interface and visualization). The MEMEX program runs at a frantic pace (interview and demos in CBS 60 minutes, briefings and demos in the White House situation room, deployment to law enforcement in February 2015). The talk will cover the goals and key challenges of the project, and describe the system we built in several domains, including the human trafficking domain with over 50 million escort ads updated hourly, and the deployment to users in law enforcement.
Biography: Dr. Pedro Szekely is a Research Team Leader at the USC Information Sciences Institute (ISI) and a Research Associate Professor at the USC Computer Science Department. Dr. Szekely joined USC in 1988 after receiving his M.S. and Ph.D. degrees in Computer Science from Carnegie Mellon University in 1982 and 1987 respectively. His research interests include Big-Data, Semantic Web and Human-Computer Interaction. His focus is on techniques and tools to extract and integrate data from a wide variety of sources (Web pages, databases, spreadsheets, etc.), and on methods to index the integrated data to support accurate querying and sophisticated analysis. The resulting software tools, Karma and DIG, released as Open Source, have been used in a variety of applications, including intelligence analysis, bioinformatics, environmental engineering and cultural heritage. A notable example is the work with the Smithsonian American Art Museum to publish the meta-data about the museums collection as Linked Open Data. Dr. Szekely is currently applying this work to combat human trafficking, deploying the tools to victim-support agencies and law enforcement.
Host: Craig Knoblock
Location: Information Science Institute (ISI) - 1135 - 11th fl 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-Using Highways for Bounded-Suboptimal Multi-Agent Path Finding
Fri, Oct 09, 2015 @ 03:00 PM - 04:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Liron Cohen, USC
Talk Title: Using Highways for Bounded-Suboptimal Multi-Agent Path Finding
Series: Natural Language Seminar
Abstract: Multi-agent path-finding (MAPF) is important for applications such as the kind of warehousing done by Kiva systems. Solving the problem optimally is NP-hard, yet finding low-cost solutions is important. Bounded-suboptimal MAPF algorithms, such as enhanced conflict-based search (ECBS), often do not perform well in warehousing domains with many agents. We therefore develop bounded-suboptimal MAPF algorithms, called CBS+HWY and ECBS+HWY, that exploit the problem structure of a given MAPF instance by finding paths for the agents that include edges from user-provided highways, which encourages a global behavior of the agents that avoids collisions. On the theoretical side, we develop a simple approach that uses highways for MAPF and provides suboptimality guarantees. On the experimental side, we demonstrate that ECBS+HWY can decrease the runtimes and solution costs of ECBS in Kiva-like domains with many agents if the highways capture the problem structures well.
Biography: Liron received a B.S. in Computer Engineering in 2007 and an M.S. in Computer Science in 2012, both from the Hebrew University of Jerusalem. Liron is interested in combinatorial problems related to constraint-based reasoning and symbolic planning. Specifically, he is looking at novel algorithmic techniques for exploiting structure in such combinatorial problems.
Host: Nima Pourdamghani and Kevin Knight
More Info: http://nlg.isi.edu/nl-seminar/
Location: 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. -
AI SEMINAR
Fri, Oct 23, 2015 @ 11:00 AM - 12:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Mohsen Taheriyan, Ph.D at USC
Talk Title: Learning the Semantics of Structured Data Sources
Series: AI Seminar
Abstract: Information sources such as relational databases, spreadsheets, XML, JSON, and Web APIs contain a tremendous amount of structured data, however, they rarely provide a semantic model to describe their contents. Semantic models of data sources capture the intended meaning of data sources by mapping them to the concepts and relationships defined by a domain ontology. Such models are the key ingredients to automate many tasks such as source discovery, data integration, and publishing semantic content on the Web. Manually modeling the semantics of data sources requires significant effort and expertise, and although desirable, building these models automatically is a challenging problem. Most of the effort to automatically build semantic models is focused on labeling the data fields (source attributes) with ontology classes and/or properties, e.g., annotating the first column of a table with the class Person and the second one with the class Movie. However, a precise semantic model needs to explicitly represent the relationships between the attributes in addition to their semantic types, e.g., stating that the person is the director of the movie. Automatically constructing such precise models is a difficult task. In this talk, I present a novel approach that exploits the knowledge from a domain ontology, the semantic models of previously modeled sources, and the vast amount of data available in the Linked Open Data (LOD) cloud to automatically learn a rich semantic model for a new source. This model represents the semantics of the new source in terms of the concepts and relationships defined by the domain ontology. The approach takes into account user corrections to learn more accurate semantic models on future data sources. Our evaluation shows that our method generates expressive semantic models for data sources and services with minimal user input.
Biography: Mohsen Taheriyan is a newly graduated PhD from the University of Southern California. He worked at Information Integration Group at ISI on learning the semantics of structured data sources. His research focus is applying Semantic Web technologies and AI techniques to understand the meaning of data. He received his B.S. in Computer Engineering from University of Tehran and his M.S. in Software Engineering from Sharif University of Technology.
Webcast: http://webcasterms1.isi.edu/mediasite/Viewer/?peid=500df65b10044d08837b95ecc188eecf1dLocation: Information Science Institute (ISI) - 1135 - 11th fl Large CR
WebCast Link: http://webcasterms1.isi.edu/mediasite/Viewer/?peid=500df65b10044d08837b95ecc188eecf1d
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: Fine Grained Temporal Patterns of Online Content Consumption
Fri, Oct 23, 2015 @ 03:00 PM - 04:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Farshad Kooti, USC/ISI
Talk Title: Fine Grained Temporal Patterns of Online Content Consumption
Series: Natural Language Seminar
Abstract: Online activity is characterized by diurnal and weekly patterns, reflecting human circadian rhythms, sleep cycles, and social patterns of work and leisure. Using data from online social networking site Facebook, we uncover temporal patterns that take place at far shorter time scales. Specifically, we demonstrate fine-grained, within-session behavioral changes, where a session is defined as a period of time a user engages with Facebook before choosing to take a break. We show that over the course of a session, users spend less time consuming some types of content, such as textual posts, and preferentially consume more photos and videos. Moreover, users who spend more time engaging with Facebook have different patterns of session activity than the less-engaged users, a distinction that is already visible at the start of the session. We study activity patterns with respect to users demographic characteristics, such as age and gender, and show that age has a strong impact on within-session behavioral changes. Finally, we show that the temporal patterns we uncover help us more accurately predict the length of sessions on Facebook.
Biography: I am a third-year Computer Science PhD student at the University of Southern California USC, Information Sciences Institute ISI working under the supervision of Kristina Lerman. My main research interest is the study of large and complex datasets, especially data from online social networks, which includes the measurement and analysis of users' behavior in OSNs. I'm currently a Data Science intern at Facebook in Menlo Park. Before joining USC, I got my master's from Max Planck Institute for Software Systems MPI SWS, Germany. I worked with Krishna Gummadi as my advisor and also with Meeyoung Cha KAIST and Winter Mason Facebook during my master's. Before MPI, I got my bachelor's in Computer Engineering Software from University of Tehran, Iran.
Host: Nima Pourdamghani 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. -
AI Seminar
Fri, Oct 30, 2015 @ 11:00 AM - 12:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Majid Janzamin, UC Irvine
Talk Title: Beating the Perils of Non-convexity: Guaranteed Training of Neural Networks Using Tensor Methods
Abstract: Training neural networks is a highly non-convex problem and in general is NP-hard. Local search methods such as gradient descent get stuck in spurious local optima, especially in high dimensions. We present a novel method based on tensor decomposition that trains a two-layer neural network with guaranteed risk bounds with polynomial sample and computational complexity. We also demonstrate how unsupervised learning can help in supervised tasks. In our context, we estimate probabilistic score functions via unsupervised learning which are then employed for training neural networks using tensor methods.
Biography: Majid Janzamin is a sixth year PhD student at the EECS Dept. at UC Irvine. He received his BSc and MSc in Electrical Engineering, from Sharif University of Technology, Tehran, Iran in 2007 and 2010, respectively. He has also visited and has done internship at Microsoft research labs at New England and Silicon Valley. His research interests are in the area of large-scale machine learning and high-dimensional statistics, and probabilistic modeling. In particular, he has worked on optimization methods for learning graphical models, and tensor methods for latent variable models.
Host: Ashish Vaswani
Webcast: http://webcasterms1.isi.edu/mediasite/Viewer/?peid=07a00eec98a44b81ab87fdfd8a6368151dLocation: Information Science Institute (ISI) - 11th floor large conference room
WebCast Link: http://webcasterms1.isi.edu/mediasite/Viewer/?peid=07a00eec98a44b81ab87fdfd8a6368151d
Audiences: Everyone Is Invited
Contact: Kary LAU
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.
