All Seminars

Title: Building Cybersecurity Programs: Lessons Learned and Future Directions
Seminar: Computer Science
Speaker: Abdallah Farraj, Ph.D, P.Eng. of Pricewaterhouse Coopers
Contact: Nosayba El-Sayed, nosayba.ae@emory.edu
Date: 2019-04-19 at 10:30AM
Venue: W201
Download Flyer
Abstract:
Many organizations look to enhance their cybersecurity capabilities as the risks of adversarial cyber activities are increasingly impacting their operations. In this talk we discuss how we can build a cybersecurity program from a practical point of view. We introduce basic cybersecurity concepts, and we discuss the technology, people and process components of mature cybersecurity programs. We focus on the activities around securing industrial control systems. We discuss future challenges and opportunities to building cybersecurity capabilities, and we present some of lessons learned from this process.
Title: The Changing Nature of Computer Science and Mathematics Research at Oak Ridge National Laboratory
Seminar: Computer Science
Speaker: Barney Maccabe of Director, Oak Ridge National Laboratory
Contact: Vaidy Sunderam, VSS@emory.edu
Date: 2019-04-12 at 10:30AM
Venue: W201
Download Flyer
Abstract:
Over the past decade or so, the computer science and mathematics research programs at ORNL like many of the DOE science laboratories has focused on computation at scale.  Two trends, an increasing emphasis on data from scientific instruments and the end of Dennard scaling, are leading to a broaden of the scope for these research programs.  In this talk, I will cover some of the history that led to the relatively narrow focus on scalable high performance computing and the ways in which ORNL is approaching the broader scope, with the intent of exploring opportunities for partnerships.
Title: Human-Centered AI through Scalable Visual Data Analytics
Seminar: Computer Science
Speaker: Minsuk Kahng of Georgia Institute of Technology
Contact: Vaidy Sunderam, vss@emory.edu
Date: 2019-04-01 at 10:00AM
Venue: Planetarium - E300 MSC
Download Flyer
Abstract:
While artificial intelligence (AI) has led to major breakthroughs in many domains, understanding machine learning models remains a fundamental challenge. They are often used as "black boxes," which could be detrimental. How can we help people understand complex machine learning models, so that they can learn them more easily and use them more effectively? In this talk, I present my research that makes AI more accessible and interpretable, through a novel human-centered approach, by creating novel data visualization tools that are scalable, interactive, and easy to learn and to use. I present my work in two interrelated topics. (1) Visualization for Industry-scale Models: I present how to scale up interactive visualization tools for industry-scale deep learning models that use large datasets. I describe how the ActiVis system helps Facebook data scientists interpret deep neural network models by visually exploring activation flows. ActiVis is patent-pending, and has been deployed on Facebook's ML platform. (2) Interactive Understanding of Complex Models: I show how visualization helps novices interactively learn complex concepts of deep learning models. I describe how I developed GAN Lab, a visual education system for Generative Adversarial Networks (GANs), one of the most popular, but hard-to-understand models. GAN Lab has been open-sourced in collaboration with Google Brain and used by over 30,000 people from 140 countries. I conclude with my vision to make AI more human-centered, to promote actionability for AI, stimulate a stronger ethical AI workforce, and foster healthy impacts of AI on broader society.
Title: From Barriers to Bridges: Designing Infrastructures for Help in Online Programming Communities
Seminar: Computer Science
Speaker: Denae Ford, PhD of North Carolina State University
Contact: Vaidy Sunderam, vss@emory.edu
Date: 2019-03-07 at 10:00AM
Venue: W303
Download Flyer
Abstract:
Online programming communities, like Stack Overflow, have norms that are not obvious nor inclusive to the 50 million programmers visiting monthly. For example, many novices ask questions that go unanswered or downvoted for not conforming to unwritten community norms. In this talk, I will present my findings from two research projects: 1) a framework of challenges users face when engaging in these communities and 2) a formative-feedback design intervention based on this framework that improved participation and acclimates users to on-site cultural norms. We find that challenges, such as a fear of negative feedback and intimidating community size, can dissuade programmers, especially novices and women, from participating in the community and forging on to become experienced contributors. To determine how to increase participation, we apply theory from a guided mentorship. In our approach, we find that mentored questions are substantially improved over non-mentored questions, with average scores increasing by 50%. These results suggest how we can challenge socio-technical communities to use identity as mechanism to increase participation.
Title: Scalable Unsupervised Phenotyping using Tensor Factorization
Seminar: Computer Science
Speaker: Ioakeim Perros, MS of Georgia Institute of Technology
Contact: Vaidy Sunderam, vss@emory.edu
Date: 2019-03-06 at 10:00AM
Venue: Atwood Chemistry, Room 360
Download Flyer
Abstract:
Originally purposed to streamline documentation of care, electronic health records (EHRs) provide a massive amount of diverse and readily available data that can be used to tackle important healthcare problems. Clinical phenotyping is one of them, which refers to identifying patient subgroups sharing common clinically-meaningful characteristics. However, there are significant challenges in using EHR data to computationally tackle this problem, related to algorithmic scalability, model interpretability and the longitudinal nature of patient data. In this talk, recent developments in the area of tensor factorization will be presented which effectively tackle those challenges.
Title: Weakly-supervised Modeling of Language, Social, and Behavioral Abstractions for Microblog Political Discourse Classification
Seminar: Computer Science
Speaker: Kristen Johnson, PhD of Purdue University
Contact: Vaidy Sunderam, vss@emory.edu
Date: 2019-03-05 at 10:00AM
Venue: W303
Download Flyer
Abstract:
Social media microblogging platforms, specifically Twitter, have become highly influential and relevant to current political events. Such platforms allow politicians to communicate with the public as events are unfolding and shape public discourse on various issues. Furthermore, by selectively using framing techniques and political slogans, politicians are able to express or conceal their stances, as well as their underlying political ideologies and moral views on policies and issues. In this talk, I will present my proposals for overcoming the challenges associated with online political discourse analysis. Specifically, I will present my approach for identifying and modeling adaptable abstractions of language and behavior which can handle the ambiguous and context-independent nature of political tweets, as well as the dynamic nature of Twitter by reducing the need for expensive annotation. My works employ relational modeling of political social networks in combination with these abstractions to accurately predict and classify the ideological stances, policy frames, and moral foundations present in tweets. I will conclude the talk by discussing my future visions of porting my abstraction and modeling techniques beyond politicians to the general public, international analysis, and the study of current policy issues.
Title: Detecting human behavior from longitudinal data streams
Seminar: Computer Science
Speaker: Afsaneh Doryab, PhD of Carnegie Mellon University
Contact: Vaidy Sunderam, vss@emory.edu
Date: 2019-03-04 at 10:00AM
Venue: Planetarium - E300 MSC
Download Flyer
Abstract:
Humans today interact frequently and intensively with a wide range of computing devices. These interactions generate data streams that often offer clues as to their physical and mental states. Analyzing and interpreting these data streams helps intelligent systems to adapt and act according to users’ needs and to provide personalized services and interventions. This capability, however, introduces new technical and social challenges to be addressed. In this talk, I will describe methods to computationally model human behavior from diverse data streams to assess the state of individuals' health and wellbeing. Through a series of systems I have built, I will also describe how models of human behavior can contribute to the seamless integration of technology into people’s lives and to connect community members for opportunistic social and economic exchange.
Title: Pattern-Based Mining of Entity/Relation Structures from Massive Text
Seminar: Computer Science
Speaker: Qi Li, PhD of University of Illinois at Urbana-Champaign
Contact: Vaidy Sunderam, vss@emory.edu
Date: 2019-02-28 at 10:00AM
Venue: W303
Download Flyer
Abstract:
Majority of information nowadays is carried by massive and unstructured text, in the form of news, articles, reports, or social media messages. This poses a major research challenge on mining entity/relation structures from unstructured text. Manual curation or labeling cannot be scalable to match the rapid growth of text. Most existing information extraction approaches rely on heavy human annotations, which can be too expensive to tune and not adaptable to new domains. In this talk, I will present a pattern-based methodology that conducts information extraction from the massive corpora using existing resources with little human effort. The first component, WW-PIE, discovers meaningful textual patterns that contain the entities of interest. The second component, TruePIE, discovers high quality textual patterns for target relation types. I will demonstrate how semi-supervised methods can empower information extraction for broad applications and provide explainable results.
Title: More Than Insights: Beyond Exploratory Data Visualization
Seminar: Computer Science
Speaker: Nam Wook Kim, PhD of Harvard University
Contact: Vaidy Sunderam, vss@emory.edu
Date: 2019-02-27 at 10:00AM
Venue: Planetarium - E300 MSC
Download Flyer
Abstract:
"The purpose of computing is insight, not numbers,” R.W. Hamming said, the founder of the ACM. This has certainly been the driving force for most visualization systems to date, which focus on exploring data to discover the unknown. However, these systems typically have complex designs that are unintuitive and cumbersome for non-expert users. On the other hand, visualizations are more and more being used to communicate data and messages to a general audience. But visualization tools for communication are still in their infancy. In this talk, I will reexamine the role of visualization beyond data exploration. I will use concrete examples from my own work to illustrate how we might go beyond traditional charts to design expressive data graphics for communication, use elements of storytelling to convey messages more effectively, and understand cognitive processes of visualization. I will conclude with my research vision for the democratization of data and speculate future research directions on deepening our understanding of visualization and designing better systems for interacting with data.
Title: Learning from Spatial-Temporal-Networked Data: Dynamics Modeling, Representation Learning, and Applications
Seminar: Computer Science
Speaker: Yanjie Fu, PhD of Missouri University of Science and Technology, University of Missouri-Rolla
Contact: Vaidy Sunderam, vss@emory.edu
Date: 2019-02-26 at 10:00AM
Venue: W303
Download Flyer
Abstract:
In this talk, I will first introduce what are spatial-temporal-networked data and why it is difficult to make sense of spatial-temporal-networked data. Then, I will focus on the modeling, representation, and applications of spatial-temporal-networked data, including (1) modeling spatial-temporal-networked **dynamics; (2) learning deep representations of spatial-temporal-networked behaviors; (3) their applications to smart transportation systems and mobile user profiling. Finally, I will conclude the talk and present the big picture on developing close-looped intelligent and trustworthy data science systems.