COMPUTER SCIENCE DIVISION UPDATE
PRESENTED BY DR. BRIAN MALLOYSEPTEMBER 23, 2016
Brian DeanWayne Goddard Sandra
Hedetmiemi
Feng LuoBrian Malloy Jim Martin
John McGregor
Murali SitaramanMark Smotherman
Pradip SrimaniJacob Sorber James Wang
Ilya Safro
Rong GeAmy Apon,
ChairHongxin Hu
CS has 16 T/TT faculty
• Jim Bottum• Jim Clements• Jill Gemmill• Alex Herzog
• Linh Ngo• Steve Stuart• Paul Wilson
Associate/Adjunct faculty
SoftwareEngineering
Data Science
Systems and Networks
Security
2016 Hiring Emphasis Areas
Energy efficient high performance computing –Rong Ge
Wearable Security and Privacy – Jacob Sorber
Research by New /Junior Faculty (Partial list only!!)
Security in SDN, security and privacy in social networks – Hongxin Hu
Fast, scalable algorithms and methods for data analytics – Ilya Safro
New CS Grants and Gifts• Amy Apon, PI, "EAGER: Characterization, Prediction, and Management of Latency Variation in Virtualized HPC and Warehouse Scale Systems", NSF, $300,000, 2016‐2018
• Amy Apon, PI, Gifts from BMW, LexisNexis, Omnibond, and Amazon, $160K total
• Hongxin Hu, PI, "Collaborative Research: CICI: Secure and Resilient Architecture: SciGuard: Building a Security Architecture for Science DMZ based on SDN and NFV Technologies", NSF, $499,805, 2017‐2019.
• Hongxin Hu, PI, Feng Luo, co‐PI, "EAGER: Defending Against Visual Cyberbullying Attacks in Emerging Mobile Social Networks", NSF, $239,680, 2015‐2017.
• Hongxin Hu, co‐PI, "Collaborative Research: CICI: Regional: SouthEast SciEntificCybersecurity for University Research (SouthEast SECURE)", NSF, $140,504, 2016.
• Hongxin Hu, co‐PI, "Development of a Security Platform for Vehicle to Infrastructure Network", Southeastern Transportation Center (University of Tennessee), $77,287, 2016‐2017.
• Hongxin Hu, co‐PI, "Alleviating the Negative Consequences of Habitual Trust on Collaborative Privacy Control in Online Social Networks", Clemson OYA Grant, $20,000, 2016.
• Hongxin Hu, PI, "III: Small: Collaborative Research: Privacy‐aware Collaborative Data Sharing in Human‐centered Social Networks", NSF, $299,767, 2015‐2018.
• Jim Martin, PI, “Extreme Orange: Engaging Clemson Students in Systems of Discovery”, IBM Faculty Award, $100,000 2016 ‐2017.
• Jim Martin, co‐PI, “Security of Distributed Cyber‐Physical Systems with Connected Vehicle Applications”, NSF, $998,581, 2015 ‐2018.
• Jim Martin, PI, Hongxin Hu, co‐PI, “Enabling Connected Vehicle Applications through Advanced Network Infrastructure”, NSF, $598,582, 2015 – 2017.
New CS Grants and Gifts, continued• Ilya Safro, PI, Gift from BMW, $40,000, 2016
• Ilya Safro, PI, "Fast and Scalable Multigrid Methods for Hypergraph Partitioning Problems", NSF, $180K, 7/2015‐8/2018
• Ilya Safro, PI "EAGER: Feedback‐based Network Optimization for Smart Cities", NSF, $151K, 8/2016‐7/2018
• Ilya Safro, co‐PI "CRISP Type 1: Data‐driven Real‐time Simulation for Adaptive Control of Interdependent Infrastructure Systems", NSF, $500K, 1/2017‐1/2019
• Murali Sitaraman, PI, Michelle P. Cook (with Florida Atlantic University and Indiana University) Collaborative Research: IUSE: Engaging and Enabling Learners to Reason Logically about Code, NSF, $300K (CU $213K), 2016‐18.
• Murali Sitaraman, PI, Russ Marion, Stacy M. Che, Michelle P. Cook, Eileen T. Kraemer, IUSE: Understanding and Propagating the Essence of Successful Computing Education Projects, NSF, $41,280, 8/1/2016 to 7/31/2017
• Jacob Sorber, PI, “Smart earpiece for supporting healthy eating behaviors”, NSF (Recommended), $1,400,000(Clemson $700,711), 9/2016
• Jacob Sorber, PI, “Computational Jewelry for Mobile Health (ABR)”, NSF (CNS‐1619950), $1,175,419 (Clemson $396,420), 9/2015
• Jacob Sorber, PI, “CAREER: Enabling Low‐Cost, Maintenance‐Free, Long‐Term Embedded Sensing Without Batteries”, NSF (CNS‐1453607), $496,661, 9/2015
Awards
• Brian Dean won the Philip Prince Award for Innovation in Teaching– only one award given at the University level each year!
• Brian Dean led the U.S. student team to the International Olympiad in Informatics (IOI), the world’s most prestigious computing contest at the high‐school level. The team earned 1 silver and 3 gold medals, a feat matched only by China and Russia.
Computer Science Research
Data Intensive Computing Ecosystems Lab within CSAVI ‐ Complex Systems, Analytics and Visualization InstituteCo‐Directors Amy Apon (CS) and Mashrur Chowdhury (CE/ITS). Some projects include:• Dynamic HPC Clusters within Amazon Web Services (AWS)
• Performance Considerations of Virtualization with Containers
• Distributed Data Delivery in a Connected Vehicle Environment
• Scalable Topic Modeling
Some of our partners:
Amy Apon
Data Analytics
Brian DeanProblem: Epileptic seizures affect more than 50 million people every year. 1 in every 150 is autistic.Our approach: Seizure localization and anomaly detection with signal processing, machine learning. Graph theory and network analysis techniques study autism.
Can we make senseof massive data sets?
Approximate quantiles
Feature selection based
on linear programming
Connectivity data structures for
geometric networks
Recommendation systems based on auctions and stable matching algorithms
Computer vision and
sparse reconstruction algorithms to build cellular mechanical
models
AlgorithmsWayne GoddardResearch in graph theory including:• Colorings• Domination• Generalized secretary problem• Eccentricity counts and sequences• Identifying codes
Managing Editor of Discrete MathematicsCo-organizer of Clemson Mini-Conference
Network and System SecurityHongxin Hu
Goals: Protecting emerging computer and network systems
Research Areas: Emerging Network Security Social Network Security Cyberbullying Defense Internet of Things Security Mobile Platform Security
Ongoing NSF-Sponsored Research Projects: SciGuard: Building a Security Architecture for Science DMZ based on SDN and NFV
Technologies ($500K) Defending Against Visual Cyberbullying Attacks in Emerging Mobile Social Networks ($258K) Privacy-aware Collaborative Data Sharing in Human-centered Social Networks ($308K) SouthEast SciEntific Cybersecurity for University Research (SouthEast SECURE) ($141K) Enabling Connected Vehicle Applications through Advanced Network Technology ($600K)
Our research explores novel systems and software design toimprove performance, scalability, and energy efficiency ofcompute- and data-intensive applications on parallel anddistributed systems.
Scalable Computing and Analytics Laboratory (ScaLab)
Scalable Computing &Analytics
Profile
Model
Control
Prototype
ApplicationsSoftware and Runtime
Multicore-/Manycore-based Clusters
TheoryEnabling
Technology
An Experimental System Approach
[email protected] web: http://www.cs.clemson.edu/~rge
Scale-Up, Scale-Out Systems
MulticoreManycore
Sponsored by
Rong Ge
• Research Projects:• Data Intensive Bioinformatics
• Large scale genome data analytics.• Precision medicine.
• Big data analytics on HPC (high performance computing) cluster
• Satellite data analytics• Machine learning method for cyberbully detection.
Feng LuoBig Data Analytics Group
• Research Support:– Defending Against Visual Cyberbullying Attacks in
Emerging Mobile Social Networks, NSF;– Development of Spark Based Software for Kratos,
Kratos Technology and Training Solutions;– Understanding contributions of transcriptome and
metabolome to autism risk: Identification of long noncoding RNA (lncRNA) and metoabolomebiomarkers and therapeutic targets for ASDs,
Self Regional Healthcare foundation.
Grammar Convergence & Program Analysis
Brian Malloy
Problem: How to reverse engineer a program to a graphical representation of the program that facilitates comprehension, analysis, testing, and debugging.
Approach: • Develop a parsable grammar• Parse the program building an
Abstract Semantic Graph (ASG)
• An ASG contains all important information about a program
• Convert the ASG into UML class diagram, control flow graph, or relevant program representation
The Internet is changing….• Converging with 75+
years of broadcast video• Approaching 4 Billion
users!• Transitioning from the
baby boomers to millennials who have transformed the Internet with Social Media
• Open data exchanges and the requirements of an ‘on-demand’ economy will redefine the scale of the Internet
Smart Infrastructure Jim Martin
Our research helps the Internet to evolve and to support emerging applications such as the Internet-of-Things. Research topics include:● Broadband access MAC enhancements and evaluation for
cable and wireless networks.● Heterogeneous wireless system architecture, protocol
development and evaluation. ● System issues related to cybersecurity, scalability, and
incremental deployability.
Vehicular and transportation Smart Grid Environmental: Mobile InternetHealth Care
Intelligent Transportation System (ITS)
Access Networks & InternetCommercial Cloud Services
Power Company’s Smart Grid Support
Cellular Operators –Licensed Spectrum
Wireless Over Unlicensed Spectrum
Smart Infrastructure
‘Coexisting Systems’
Consumer Internet (Internet Video, …)
Building Robust SoftwareProblem: Our health, wealth, and livelihood depend upon software doing its job correctly and doing it when required.Our research program includes: • Ecosystem perspectives on strategic software engineering
• Product line manufacturing of software‐intensive products
• Techniques for detailed design of fault and error propagation to ensure that abnormal behavior is handled before a failure occurs
• Techniques for evaluating and tracking the value of reusable assets by measuring deviations from the desired (technical debt)
John D. McGregor
Solving Some of Society’s Most Pressing Problems
Ilya Safro
Research Interests
• Scientific Computing• Network Science• Optimization• Machine Learning• Data Mining• Complex Systems• Graph Algorithms
Success Story I: Multiscale Methods for Fast Data MiningClassification, Detection of Outliers, Clustering
Algebraic Multigrid Framework Excellent quality of performance measures Up to 1000 times faster on
a single processor! Scalable
Attacked areas Optimal areas to prevent the spread
Fast and scalable solver for computationally hard optimization problem
High‐quality model for optimalresponse
Budget efficient optimization results
Success Story II: Response to Cyber Attacks, Epidemics and Information Spread on Networks
Software Engineering
Problem: Verify correctness of softwareSupport: NSF Funding continuously for about 25 years; Current support through NSF Software-Hardware Foundations, REU Supplements.
Problem: Understanding learning obstacles and teaching analytical ReasoningApproach: Web-integrated toolsDissemination: Multiple educator workshops, A dozen adoptee institutionsSupport: Current support through NSF IUSE Collaborators: College of Education at Clemson; others
Teaching Reasoning
Kimberly Ferguson, Christin Marshall, John McCann
RESOLVE Verifying Compiler
Murali Sitaraman
Mobile Sensing for the Real World
Jacob Sorber
Problems in ecology such as wildlife tracking, Intelligent River, or with chronic disease or addiction research require advanced monitoring technologies
Approach• Mobile sensing for the real world – uncertain
conditions + scarce resources• Energy (harvesting and management) –
Conserve, Adapt, and Emulate• RFID-scale power — failure as the common case• Security & Privacy – passive authentication, safe
computing
Problem: Bioinformatics researchers need a way to share code and data
Our Approach:• G-SESAME tools have been
used more than 25.2 million times by researchers from 65 countries
• The G-BEAN search engine is shown to be better than PubMed, the most used search interface for biomedical articles
• “A New Method to Measure the Semantic Similarity of GO Terms”, Bioinformatics, 2007, 23: 1274-1281; cited more than 200 times in last 5 years
Having Impact
http://bioinformatics.clemson.edu/G-SESAME/
James Wang Pradip SrimaniMultimedia and Informatics
James Wang Pradip Srimani
• Information Retrieval• Biomedical Image Processing• Analysis and Classification• Wireless and Mobile Networks• Cloud computing, P2P• Multimedia
New NIH grant: “Methods for Retrospective Multi‐site Research”, NIH 1R01HD069374‐01A1, (PI: Mark Eckert, MUSC), $1,338,555, (05/12 ~ 05/17)Three continuing NSF grants, $1.3M
Having ImpactMultimedia and Informatics
COMPUTER SCIENCE DIVISION