Select a calendar:
Filter March Events by Event Type:
SUNMONTUEWEDTHUFRISAT
Conferences, Lectures, & Seminars
Events for March
-
CS Colloquium: Rahul Chatterjee (Cornell University) - Empiricism-Informed Secure System Design: From Improving Passwords to Helping Domestic Violence Victims
Mon, Mar 04, 2019 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Rahul Chatterjee, Cornell University
Talk Title: Empiricism-Informed Secure System Design: From Improving Passwords to Helping Domestic Violence Victims
Series: CS Colloquium
Abstract: Security often fails in practice due to a lack of understanding of the nuances in real-world systems. For example, users choose weak passwords to deal with the several usability issues with passwords, which in turn degrades the security of passwords. I will talk about how we can build better security mechanisms by combining methodical empiricism with analytical frameworks. First, in the context of passwords, I will show how to improve the usability of passwords by allowing users to log in with typos in their passwords. I will detail in the talk how to do so without giving attackers any additional advantage to impersonate a user.
In the second part of my talk, I will talk about my recent research direction on how traditional authentication mechanisms fail to properly model digital attacks by domestic abusers, and therefore are ineffective for victims. As a result, abusers can spy on, stalk, or harass victims using seemingly innocuous apps and technologies. I will finish with some recent progress that I have made in helping victims of tech abuse, and provide some future research directions.
This lecture satisfies requirements for CSCI 591: Research Colloquium.
Biography: Rahul Chatterjee is a Ph.D. candidate at Cornell University, working on computer security. Prior to joining Cornell, Rahul received his masters from the University of Wisconsin-Madison and bachelors from the Indian Institute of Technology (IIT), Kharagpur. Rahul's research focuses on user authentication, in particular passwords and biometrics. Lately, he is also conducting research on how to help stop technology abuse in the context of domestic violence. His co-authored papers have been covered by several media outlets, including The New York Times, and the MIT Tech Review. His work on password typos was recognized with the distinguished student paper award at IEEE S&P (2016).
Host: Muhammad Naveed
Location: Ronald Tutor Hall of Engineering (RTH) - 115
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Tatsunori Hashimoto (Stanford University) - Beyond the average case: machine learning for atypical examples
Tue, Mar 05, 2019 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Tatsunori Hashimoto, Stanford University
Talk Title: Beyond the average case: machine learning for atypical examples
Series: CS Colloquium
Abstract: Although machine learning systems have improved dramatically over the last decade, it has been widely observed that even the best systems fail on atypical examples. For example, prediction models such as image classifiers have low accuracy on images from minority cultures, and generative models such as dialogue systems are often incapable of generating diverse, atypical responses. In this talk, I will discuss two domains where high performance on typical examples is insufficient.
The first is learning prediction models that perform well on minority groups, such as non-native English speakers using a speech recognition system. We demonstrate that models with low average loss can still assign high losses to minority groups, and this gap can amplify over time as minority users that suffer high losses stop using the model. We develop an approach using distributionally robust optimization that learns models that perform well over all groups and mitigate the feedback loop.
The second domain is learning natural language generation (NLG) systems, such as a dialogue system. It has been frequently observed that existing NLG systems which produce high-quality samples rely heavily on typical responses such as "I don't know" and fail to generate the full diversity of atypical but valid human responses.
We carefully quantify this problem through a new evaluation metric based on the optimal classification error between human- and model-generated text and propose a new, edit-based generative model of text whose outputs are both diverse and high-quality.
This lecture satisfies requirements for CSCI 591: Research Colloquium.
Biography: Tatsunori (Tatsu) Hashimoto is a 3rd year post-doc in the Statistics and Computer Science departments at Stanford, supervised by Professors Percy Liang and John Duchi. He holds a Ph.D from MIT where he studied random walks and computational biology under Professors Tommi Jaakkola and David Gifford, and a B.S. from Harvard in Statistics and Math. His work has been recognized in NeurIPS 2018 (Oral), ICML 2018 (Best paper runner-up), and NeurIPS 2014 Workshop on Networks (Best student paper).
Host: Yan Liu
Location: Olin Hall of Engineering (OHE) - 132
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Behnam Neyshabur (New York University) - Why Do Neural Networks Learn?
Wed, Mar 06, 2019 @ 09:00 AM - 10:00 AM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Behnam Neyshabur, New York University
Talk Title: Why Do Neural Networks Learn?
Series: CS Colloquium
Abstract: Neural networks used in practice have millions of parameters and yet they generalize well even when they are trained on small datasets. While there exists networks with zero training error and a large test error, the optimization algorithms used in practice magically find the networks that generalizes well to test data. How can we characterize such networks? What are the properties of networks that generalize well? How do these properties ensure generalization?
In this talk, we will develop techniques to understand generalization in neural networks. Towards the end, I will show how this understanding can help us design architectures and optimization algorithms with better generalization performance.
This lecture satisfies requirements for CSCI 591: Research Colloquium.
Biography: Behnam Neyshabur is a postdoctoral researcher in Yann LeCun's group at New York University. Before that, he was a member of Theoretical Machine Learning program lead by Sanjeev Arora at the Institute for Advanced Study (IAS) in Princeton. In summer 2017, he received a PhD in computer science at TTI-Chicago where Nati Srebro was his advisor. He is interested in machine learning and optimization and his primary research is on optimization and generalization in deep learning.
Host: Haipeng Luo
Location: Ronald Tutor Hall of Engineering (RTH) - 109
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium:Sida Wang (Princeton University) - Learning Adaptive Language Interfaces Through Interaction
Wed, Mar 06, 2019 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Sida Wang, Princeton University
Talk Title: Learning Adaptive Language Interfaces Through Interaction
Series: CS Colloquium
Abstract: The interactivity and adaptivity of natural language have the potential to allow people to better communicate with increasingly AI-driven computer systems. However, current natural language interfaces are mostly static and fall short of their potential. In this talk, I will cover two systems that can quickly learn from interactions, adapt to users, and simultaneously give feedback so that users can adapt to the system. The first system learns from scratch from users in real time. The second starts with a programming language and then learns to naturalize the programming language by interacting with users. Finally, I will present how these ideas can be combined to build a natural language interface for data visualization and discuss my work on modeling interactive language learning more rigorously.
This lecture satisfies requirements for CSCI 591: Research Colloquium.
Biography: Sida Wang is a research instructor at Princeton University and Institute for Advanced Study working in the areas of natural language processing and machine learning. He holds a Ph.D. in computer science from Stanford University and a B.A.Sc. from the University of Toronto. He received an outstanding paper award at ACL 2016 and the NSERC Postgraduate Scholarship.
Host: Joseph Lim
Location: Ronald Tutor Hall of Engineering (RTH) - 115
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CAIS Seminar: Lindsay Young (University of Chicago) - Social Network Analysis and Artificial Intelligence: Methodological Partners in the Study of HIV Prevention and Risk Online
Wed, Mar 06, 2019 @ 04:00 PM - 05:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Dr. Lindsay Young, University of Chicago
Talk Title: Social Network Analysis and Artificial Intelligence: Methodological Partners in the Study of HIV Prevention and Risk Online
Series: USC Center for Artificial Intelligence in Society (CAIS) Seminar Series
Abstract: As transmitters of information and progenitors of behavioral norms, social networks are critical mechanisms of HIV prevention and risk in impacted populations like men who have sex with men (MSM), people who inject drugs (PWID), and homeless youth. Today, widespread use of online social networking technologies (e.g., Facebook, Instagram, Twitter) yield unprecedented amounts of relational and communication data far richer than anything previously collected in offline (physical) network settings. However, parsing these complex data into tractable insights and solutions requires an innovative and flexible computational toolkit that extends beyond traditional approaches. In this talk, Dr. Young will discuss her ongoing efforts to unpack how HIV prevention and risk manifest in the Facebook networks of young MSM using a hybrid of computational methods that include social and semantic network analysis and machine learning approaches for textual analysis and predictive modeling. She will conclude with a discussion of the practical implications of this work and outstanding challenges that require further exploration.
This lecture satisfies requirements for CSCI 591: Research Colloquium.
Biography: Dr. Lindsay Young is a NIH Pathway to Independence Award Postdoctoral Fellow at the University of Chicago Department of Medicine and Chicago Center for HIV Elimination (CCHE). Trained as a social scientist and network methodologist, she now applies those perspectives to understand the social and communicative contexts of HIV risk and prevention among young sexual minorities and other vulnerable populations. She is particularly interested in how online social network data can be leveraged for behavioral research and interventions.
Host: Milind Tambe
Location: James H. Zumberge Hall Of Science (ZHS) - 252
Audiences: Everyone Is Invited
Contact: Computer Science Department
-
CS Colloquium: Eunsol Choi (University of Washington) - Learning to Understand Entities In Text
Thu, Mar 07, 2019 @ 09:30 AM - 10:30 AM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Eunsol Choi, University of Washington
Talk Title: Learning to Understand Entities In Text
Series: CS Colloquium
Abstract: Real world entities such as people, organizations and countries play a critical role in text. Reading offers rich explicit and implicit information about these entities, such as the categories they belong to, relationships they have with other entities, and events they participate in. In this talk, we introduce approaches to infer implied information about entities, and to automatically query such information in an interactive setting. We expand the scope of information that can be learned from text for a range of tasks, including sentiment extraction, entity typing and question answering. To this end, we introduce new ideas for how to find effective training data, including crowdsourcing and large-scale naturally occurring weak supervision data. We also describe new computational models, that represent rich social and conversation contexts to tackle these tasks. Together, these advances significantly expand the scope of information that can be incorporated into the next generation of machine reading systems.
This lecture satisfies requirements for CSCI 591: Research Colloquium.
Biography: Eunsol Choi is a Ph.D candidate at the Paul G. Allen School of Computer Science at the University of Washington. Her research focuses on natural language processing, specifically applying machine learning to recover semantics from text. She completed a B.A. in Computer Science and Mathematics at Cornell University, and is a recipient of the Facebook fellowship.
Host: Xiang Ren
Location: Ronald Tutor Hall of Engineering (RTH) - 109
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Chi Jin (UC Berkeley) Machine Learning: Why Do Simple Algorithms Work So Well?
Thu, Mar 07, 2019 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Chi Jin, UC Berkely
Talk Title: Machine Learning: Why Do Simple Algorithms Work So Well?
Series: CS Colloquium
Abstract: While state-of-the-art machine learning models are deep, large-scale, sequential and highly nonconvex, the backbone of modern learning algorithms are simple algorithms such as stochastic gradient descent, or Q-learning (in the case of reinforcement learning tasks). A basic question endures---why do simple algorithms work so well even in these challenging settings?
This talk focuses on two fundamental problems: (1) in nonconvex optimization, can gradient descent escape saddle points efficiently? (2) in reinforcement learning, is Q-learning sample efficient? We will provide the first line of provably positive answers to both questions. In particular, we will show that simple modifications to these classical algorithms guarantee significantly better properties, which explains the underlying mechanisms behind their favorable performance in practice.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Chi Jin is a Ph.D. candidate in Computer Science at UC Berkeley, advised by Michael I. Jordan. He received a B.S. in Physics from Peking University. His research interests lie in machine learning, statistics, and optimization, with his PhD work primarily focused on nonconvex optimization and reinforcement learning.
Host: Haipeng Luo
Location: Olin Hall of Engineering (OHE) - 132
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Gang Wang (Virginia Tech) - Human Augmentation for Internet Security
Mon, Mar 18, 2019 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Gang Wang, Virginia Tech
Talk Title: Human Augmentation for Internet Security
Series: CS Colloquium
Abstract: Human factors are playing a critical role in the security of today's Internet systems. On one hand, human factors are constantly exploited by attackers to launch serious attacks, leading to massive data breaches and ransomware infections. On the other hand, human (expert) intelligence is instrumental in detecting and combating new threats (e.g., zero-days) that automated methods such as machine learning often fail to capture.
In this talk, I will describe our efforts to improve security through human augmentation. Human augmentation includes (1) reducing the security risks introduced by human factors, and (2) integrating human intelligence to build more robust security defenses. First, I will describe our progress to reduce the risk of human factors by detecting and mitigating flawed system designs that severely weaken user-level defenses. Using spear phishing as an example, I will illustrate how data analytics and active measurements can make a key difference in this process. Second, I will share our recent results on improving the trust and robustness of security systems by generating "human-interpretable" outputs. By building an "explanation system" for deep learning based security applications, we allow security analysts to diagnose classification errors and patch model weaknesses. Finally, I conclude by highlighting my future plans of using data-driven approaches to augmenting security defenses for both humans and algorithms.
This lecture satisfies requirements for CSCI 591: Research Colloquium.
Biography: Gang Wang is an Assistant Professor of Computer Science at Virginia Tech. He obtained his Ph.D. from UC Santa Barbara in 2016, and a B.E. from Tsinghua University in 2010. His research focuses on human (user) aspects of Internet security. His work takes a data-driven approach to addressing emerging security threats in massive communication systems (social networks, email services), crowdsourcing systems, mobile applications, and enterprise networks. He is a recipient of the NSF CAREER Award (2018), Google Faculty Research Award (2017), ACM CCS Outstanding Paper Award (2018), and SIGMETRICS Best Practical Paper Award (2013). His research has appeared in a diverse set of top-tier venues in Security, Measurement, Networking, and HCI. His projects have been covered by media outlets such as MIT Technology Review, The New York Times, Boston Globe, CNN, ACM TechNews, and New Scientist.
Host: Aleksandra Korolova
Location: Ronald Tutor Hall of Engineering (RTH) - 115
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: ShiQing Ma (Purdue University) - Transparent Computing Systems Enabled by Program Analysis
Tue, Mar 19, 2019 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: ShiQing Ma, Purdue University
Talk Title: Transparent Computing Systems Enabled by Program Analysis
Series: CS Colloquium
Abstract: Modern computing systems are complex and opaque, which is the root cause of many security and software engineering problems. In enterprise level system operations, this leads to inaccurate and hard-to-understand attack forensics results. In deep learning systems, such opaqueness prevents us from understanding the misclassifications and improving the model accuracy. Hence, there is a pressing need for improving the transparency of these systems to help us solve the corresponding security and software engineering problems.
In this talk, I will focus on my research efforts of developing novel program analysis techniques to improve the transparency of such systems and their applications in attack forensics and deep learning systems. For attack forensics, I will first describe a compiler-based execution partitioning technique MPI which helps accomplish accurate, semantics-rich and multi-perspective attack forensics. For deep learning systems, I will introduce novel state differential analysis and input selection techniques to analyze deep learning model internals for addressing the misclassification problem. Finally, I will briefly present my ongoing and future work on intelligent systems (i.e., systems that combine traditional computing components and artificial intelligent components).
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Shiqing Ma is a Ph.D. candidate in the Department of Computer Science at Purdue University, co-advised by Professors Xiangyu Zhang and Dongyan Xu. His research interests lie in solving security and software engineering problems via program analysis techniques with a focus on improving the transparency of modern computing systems. He is the recipient of two Distinguished Paper Awards at ISOC NDSS 2016 and USENIX Security 2017
Host: Muhammad Naveed
Location: Olin Hall of Engineering (OHE) - 132
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Protiva Rahman (Ohio State University) - Amplifying Domain Expertise in Data Pipelines
Tue, Mar 19, 2019 @ 04:00 PM - 05:20 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Protiva Rahman, Ohio State University
Talk Title: Amplifying Domain Expertise in Data Pipelines
Series: Computer Science Colloquium
Abstract: Digitization of forms and electronic health records (EHR) has made data from diverse domains available for analysis. The specialized nature of the data require domain expert input at every step of the data analysis pipeline, including entry, cleaning, and analysis. Since domain experts (e.g. physicians) are highly skilled in their fields, their time is very valuable and expensive. Moreover, they often do not have any training in computer science or statistics, making it difficult for them to effectively interact with data. Thus, it is crucial that we make data interaction easy, efficient and effortless for experts. This involves amplifying or generalizing their inputs to multiple data points, reducing their time and effort.
In this talk, I will present Icarus, a system that leverages the database schema to amplify domain expert input during data cleaning. Icarus optimizes a weighted sum to guide the user to high-impact edits. Once a user fills in a cell, the system leverages the many-to-one relations in the database to suggest generalized update queries in the form of rules. These rules apply to a larger number of cells, amplifying the user's single edit.
This lecture satisfies requirements for CSCI 591: Research Colloquium.
Biography: Protiva Rahman is a fourth-year Ph.D. student in the Department of Computer Science and Engineering at the Ohio State University, advised by Professor Arnab Nandi. Her research interests include databases, human-computer interaction, visualization, and clinical informatics. Besides data cleaning, she has also worked on optimizing data entry interfaces for constrained interaction, guidelines for evaluating interactive systems and visualizations for domain expert consensus.
Host: Computer Science Department
Location: Henry Salvatori Computer Science Center (SAL) - 101
Audiences: Everyone Is Invited
Contact: Computer Science Department
-
CAIS Seminar: Robin Petering, Nick Barr, & AJ Srivastava - MyPath: Intervention for Reduction in Violence among at Risk Youth
Wed, Mar 20, 2019 @ 04:00 PM - 05:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Dr. Robin Petering, Dr. Nick Barr, & Dr. AJ Srivastava,
Talk Title: MyPath: Intervention for Reduction in Violence among at Risk Youth
Series: USC Center for Artificial Intelligence in Society (CAIS) Seminar Series
Abstract: One in ten young people between ages 18-25 experience homelessness every year in the United States. The lives of youth experiencing homelessness is characterized by violence, more so than their housed counterparts. This is the result of several, often co-occurring, risk factors such as experience of childhood trauma, subsistence survival strategies including drug and alcohol use, and exposure to perpetrators during street tenure. Violence has many consequences from physical injury to heightened mental health distress. Reducing exposure and engagement to violence is critical for safe and successful exit from homelessness. However, to date, implementing a violence reduction intervention in this community has had little success.
MyPath is a peer-based social network intervention designed to reduce experiences of violence in a community of young persons experiencing homelessness. MyPath utilizes strategic machine learning selection methods to identify potential Mindfulness and Yoga Peer Ambassadors and invites them to participate in an intensive 3-hour mindfulness and yoga retreat that relates the two practices to the impact of violence. The retreat is followed by weekly 1-hour trainer-facilitated mindfulness and yoga classes that are open for attendance of non-peer ambassadors as well. MyPath is novel in that it uses an algorithm, called ViolMin, to identify potential peer ambassadors, which takes into account the uncertainty in links of surveyed network data and identifies "influential" individuals in a network, those who have a history of violent behavior, and yet are open to intervention.
The MyPath Pilot was implemented in partnership with Safe Place for Youth (SPY) during the summer of 2018. During this project, eight Mindfulness and Yoga Peer Ambassadors, selected by the ViolMin algorithm, participated in the program. Six weeks after the introduction of the MyPath programming, pilot results showed a statistically significant reduction in violence. The number of young people involved in physical fights dropped by 40%. There was also an increase of 85% in number of individuals who practice regular mindfulness and yoga. Moreover, the selected Mindfulness and Yoga Peer Ambassadors were highly engaged in the program. One MyPath ambassador reflected on the program, "People think threatening and violence is the answer. If everyone did mindfulness we would be living in a semi-better world. I didn't know anything about mindfulness, all I did know was violence, how to protect myself. When I got to SPY, I learned mindfulness and learned how to relax with yoga. I feel like a different person when I do it."
Biography: Dr. Robin Petering is interested in improving the lives of young persons who experience homelessness through community-based research, policy advocacy and program implementation. Her research agenda focuses on reducing violence through innovative intervention approaches. @robinpetering
Dr. Nicholas Barr is a postdoctoral fellow interested in improving mental and behavioral health outcomes in populations at high risk for adverse experiences. His research agenda focuses on investigating the protective effects of mindfulness and emotion regulation.
Dr. Ajitesh Srivastava is a Postdoctoral Research Associate at Ming Hsieh Institute of Electrical Engineering. His research interests include Data Mining, Social Network Analysis, Graph Algorithms, Optimizations, and Parallel Computing.
Host: Milind Tambe
Location: James H. Zumberge Hall Of Science (ZHS) - 252
Audiences: Everyone Is Invited
Contact: Computer Science Department
-
CS Colloquium: Amy Zhang (MIT) - Systems to Improve Online Discussion
Thu, Mar 21, 2019 @ 09:30 AM - 10:30 AM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Amy Zhang, MIT
Talk Title: Systems to Improve Online Discussion
Series: CS Colloquium
Abstract: Discussions online are integral to everyday life, affecting how we learn, work, socialize, and participate in public society. Yet the systems that we use to conduct online discourse, whether they be email, chat, or forums, have changed little since their inception many decades ago. As more people participate and more venues for discourse migrate online, new problems have arisen, and old problems have intensified. People are still drowning in information and must now juggle dozens of disparate discussion silos in addition. Finally, an unfortunately significant proportion of this online interaction is unwanted or unpleasant, with clashing norms leading to people bickering or getting harassed into silence. My research in human-computer interaction is on reimagining outdated designs towards designing novel online discussion systems that fix what's broken about online discussion. To solve these problems, I develop tools that empower users and communities to have direct control over their experiences and information. These include: 1) summarization tools to make sense of large discussions, 2) annotation tools to situate conversations in the context of what is being discussed, as well as 3) moderation tools to give users more fine-grained control over content delivery.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Amy X. Zhang is a graduate student at MIT's Computer Science and Artificial Intelligence Laboratory, focusing on human-computer interaction and social computing, and a 2018-19 Fellow at the Harvard Berkman Klein Center. She has interned at Microsoft Research and Google Research, received awards at ACM CHI and CSCW, and featured in stories by ABC News, BBC, CBC, and more. She has an M.Phil. in CS at University of Cambridge on a Gates Fellowship and a B.S. in CS at Rutgers, where she captained the Division I Women's tennis team. Her research is supported by a Google PhD Fellowship and an NSF Graduate Research Fellowship.
Host: Nora Ayanian
Location: Ronald Tutor Hall of Engineering (RTH) - 109
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Srijan Kumar (Stanford University) - Data Science for Healthy Online Interactions
Thu, Mar 21, 2019 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Srijan Kumar, Stanford University
Talk Title: Data Science for Healthy Online Interactions
Series: CS Colloquium
Abstract: The web enables users to interact with one another and shape opinion at an unprecedented speed and scale. However, the prevalence of disinformation and malicious users makes the web unsafe and unreliable, for example, 40% of users have experienced online harassment and platforms have disabled user comments because of trolling. In this talk, I will present data science methods that help us to create a better and safer web ecosystem for everyone. In particular, I will present methods to extract knowledge from the social graph structure and augment with behavior signals to characterize, detect, and mitigate the damage of disinformation and malicious users.
First, I will describe a graph mining collective classification algorithm to identify fake reviews on e-commerce platforms. The method learns trustworthiness scores from the user-to-product review network to identify sophisticated fraudsters. The method is currently being used in production at Flipkart, India's largest e-commerce platform. Next, I will present the first web-scale characterization of multiple account abuse in online discussions and my method of statistical analysis of user interaction graphs to detect them. Finally, I will show how learning embeddings from the social network structure helps to predict online conflicts and to mitigate their damage. These methods power online tools that help administrators in Reddit and Wikipedia.
I will conclude the talk by describing my future research directions that will enable us to proactively predict how malicious behavior will evolve in the future, both on web platforms and face-to-face interactions
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Srijan Kumar (https://stanford.edu/~srijan/) is a postdoctoral scholar in Computer Science at Stanford University. His research investigates data science and machine learning to create healthy online and offline interactions, focusing on developing methods to curb deception, misbehavior, and disinformation. His methods have had a tangible real-world impact and are being used at major tech companies, including Flipkart, Reddit, and Wikipedia. His research has received the ACM SIGKDD Doctoral Dissertation Award runner-up 2018, Larry S. Davis Doctoral Dissertation Award 2018, and WWW Best Paper Award runner-up 2017. His research is interdisciplinary and has been included in the curriculum at several universities, including UIUC, University of Michigan, and Stanford University. His research has been included in documentary (Familiar Shapes) and covered in popular press, including CNN, The Wall Street Journal, Wired, and New York Magazine. He did his Ph.D. in Computer Science from University of Maryland, College Park, and B.Tech. from Indian Institute of Technology, Kharagpur.
Host: Xiang Ren
Location: Olin Hall of Engineering (OHE) - 132
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Anand Iyer (University of California, Berkeley) - Scalable Systems for Large-Scale Dynamic Connected Data Processing
Mon, Mar 25, 2019 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Anand Iyer, University of California, Berkeley
Talk Title: Scalable Systems for Large-Scale Dynamic Connected Data Processing
Series: CS Colloquium
Abstract: As the proliferation of sensors rapidly make the Internet-of-Things (IoT) a reality, the devices and sensors in this ecosystem-”such as smartphones, video cameras, home automation systems and autonomous vehicles-”constantly map out the real-world producing unprecedented amounts of connected data that captures complex and diverse relations. Unfortunately, existing big data processing and machine learning frameworks are ill-suited for analyzing such dynamic connected data, and face several challenges when employed for this purpose.
In this talk, I will present my research that focuses on building scalable systems for dynamic connected data processing. I will discuss simple abstractions that make it easy to operate on such data, efficient data structures for state management, and computation models that reduce redundant work. I will also describe how bridging theory and practice with algorithms and techniques that leverage approximation and streaming theory can significantly speed up computations. The systems I have built achieve more than an order of magnitude improvement over the state-of-the-art and are currently under evaluation in the industry for real-world deployments.
This lecture satisfies requirements for CSCI 591: Research Colloquium.
Biography: Anand Iyer is a PhD candidate at the University of California, Berkeley advised by Prof. Ion Stoica. His research interest is in systems with a current focus on enabling efficient analysis and machine learning on large-scale dynamic, connected data. He is a recipient of the Best Paper Award at SIGMOD GRADES-NDA 2018 for his work on approximate graph analytics. Before coming to Berkeley, he was a member of the Mobility, Networking and Systems group at Microsoft Research India. He completed his M.S at the University of Texas at Austin.
Host: Barath Raghavan
Location: Ronald Tutor Hall of Engineering (RTH) - 115
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: TBA
Tue, Mar 26, 2019 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: TBA, TBA
Talk Title: TBA
Series: CS Colloquium
Host: Ramesh Govindan
Location: Olin Hall of Engineering (OHE) - 132
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Abe Davis (Stanford University) - Augmenting Imagination: Capturing, Modeling, and Exploring the World Through Video
Thu, Mar 28, 2019 @ 09:30 AM - 10:30 AM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Abe Davis, Stanford University
Talk Title: Augmenting Imagination: Capturing, Modeling, and Exploring the World Through Video
Series: CS Colloquium
Abstract: Cameras offer a rich and ubiquitous source of data about the world around us, providing many opportunities to explore new computational approaches to real-world problems. In this talk, I will show how insights from art, science, and engineering can help us connect progress in visual computing with typically non-visual problems in other domains, allowing us to leverage the convenience and power of video to solve new problems. The first section of the talk will focus on visual vibration analysis: I will show how insights from physics can help us extract sound from silent video, reason about structural and material properties that are perceptually invisible to humans, and even build interactive physical simulations of visible objects. The second section of the talk will give an overview of how similar methodologies can be applied to artistic domains, using insights from music, dance, and cinematography to design computational tools that offer creative control over large amounts of media.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Abe Davis is a postdoctoral researcher at Stanford University working at the intersections of computer graphics, vision, HCI, and civil engineering. He earned his Ph.D. in Electrical Engineering and Computer Science from MIT in 2016 and is the recipient of the MIT Sprowls Award for Outstanding Dissertation in Computer Science and the ACM SIGGRAPH Outstanding Doctoral Dissertation Honorable Mention Award. Abe was awarded NSF and Mathworks graduate fellowships, named one of Forbes Magazine's "30 under 30", Business Insider's "50 Scientists Who are Changing the World" and "8 Innovative Scientists in Tech and Engineering." As a postdoc, he won the "Most Practical SHM Solution for Civil Infrastructures" Award at IWSHM 2017, and has been the recipient of two Magic Grants from the Brown Institute for Media Innovation.
Host: Jernej Barbic
Location: Ronald Tutor Hall of Engineering (RTH) - 109
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: TBA
Thu, Mar 28, 2019 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: TBA, TBA
Talk Title: TBA
Series: CS Colloquium
Abstract: TBA
This lecture satisfies requirements for CSCI 591: Research Colloquium.
Biography: TBA
Host: Ramesh Govindan
Location: Olin Hall of Engineering (OHE) - 132
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
