PhD Defense

Visualizing Sequential Patterns in Large

Datasets Using Levels of Abstraction

Dana Wortman

11am – 2pm, Friday, 1 November 2013, ITE 325b

Student retention and success are important topics in all academic fields and institutions. Faculty seek to understand which topics, theories, or skills defeat students or require strengthening to promote success. Programs seek to understand how to better sequence courses to ensure students are prepared for requisite future courses. Institutions seek to understand how to intervene to promote retention and improve graduation rates. Unfortunately, most statistics gathered by Institutional Research efforts are limited to failure rates, enrollment rates, and graduation rates and do not often explore individual student performance. While these are often further analyzed by various student demographic attributes such as race and gender, these statistical methods alone are insufficient to understand student performance over time and sequential patterns of enrollment or success and failure. This research presents a method using multiple levels of abstraction to visualize performance patterns over time.

To visualize student enrollment and performance patterns, several issues must be addressed including sequential versus concurrent enrollment, spatial layout of course events, and performance over time. Another challenge addressed by this work is that of presenting sequences within the context of the entire program. To address these issues, multiple abstractions are combined in a multi-layered visualization that presents a high-level overview of students enrollment and performance patterns while retaining detailed information regarding individual student progress and performance as they advance through their courses.

The aggregated view represents the lowest level of abstraction, student enrollment and performance are aggregated into a graph structure, presenting patterns of movement throughout the program at the individual course level. The clustered view represents mined sequential patterns of enrollment and performance, illustrating common sequences. The directed view represents the highest level of abstraction and uses two visual elements, heat maps and a vector field, to illustrate overall performance in individual events and movement through the program. Results from multiple cohorts can then be superimposed on the same visualization to enable easy comparisons between patterns. Together, these abstractions provide a focus+context view of student performance, retaining outliers and emphasizing common patterns to illuminate dominant and unique patterns between cohorts of students.

This approach can help educators better understand student progress through the program, performance in individual courses, or student-selected course sequencing and this information can be used to address deficiencies in preparation, skills, or prerequisites. To demonstrate the appropriateness of this approach, performance and enrollment patterns are explored in the Computer Science program at the University of Maryland, Baltimore County. Specifically, this work examines the Gateway policy that requires students to earn a B or higher in the first two required programming courses before progressing with the hopes of validating the existing Gateway but also exploring other possible Gateway courses. Other issues explored within the Computer Science program include race, gender, math placement, and high school scores with the goal of attracting and retaining a more diverse group of students.

Committee: Penny Rheingans (chair), Marie desJardins, Marc Olano, Tim Finin and Diane Lee

Center for Information Security and Assurance
University of Maryland, Baltimore County

Cloud Forensics and its Many Challenges

Ken Zatyko

Assured Information Security, Inc.

12-1pm, Friday 1 November 2013, ITE 229, UMBC

In this presentation, we present a challenge question for today’s cyber experts, cyber scientists, and cyber analysts. Does Locard’s Exchange Principle apply in digital forensics? The dramatic increase in cybercrime and the repeated cyber intrusions into critical infrastructure demonstrate the need for improved security. The Executive Office of the President noted on May 12, 2011 “cyber threat is one of the most serious economic and national security challenges we face as a nation.” We believe addressing whether or not Locard’s Exchange Principle applies to digital forensics is a fundamental question that can guide or limit the scientific search for digital evidence.

Locard’s Exchange Principle is often stated in forensics publications “every contact leaves a trace…” Essentially Locard’s Exchange Principle is applied to crime scenes in which the perpetrator(s) of a crime comes into contact with the scene. The perpetrator(s) will both bring something into the scene, and leave with something from the scene. In the cyber world, the perpetrator may or may not come in physical contact with the crime scene, thus, this brings a new facet to crime scene analysis. According to the World of Forensic Science, Locard’s publications make no mention of an “exchange principle,” although he did make the observation “Il est impossible au malfaiteur d’agir avec l’intensité que suppose l’action criminelle sans laisser des traces de son passage.” (It is impossible for a criminal to act, especially considering the intensity of a crime, without leaving traces of this presence.)

The term “principle of exchange” first appears in Police and Crime-Detection, in 1940, and was adapted from Locard’s observations. The field of digital forensics can be strictly defined as “the application of computer science and investigative procedures for a legal purpose involving the analysis of digital evidence after proper search authority, chain of custody, validation with mathematics, use of validated tools, repeatability, reporting, and possible expert presentation. (Zatyko, 2007).” Furthermore, digital evidence is defined as information stored or transmitted in binary form that may be relied on in court. (National Institute of Justice, 2004). However, digital forensics tools and techniques have also been used by cyber analysts and researchers to conduct media analysis, compile damage assessments, build timelines, and determine attribution. According to the Department of Defense Cyber Crime Center’s training program, cyber analysts require knowledge on how network intrusions occur, how various logs are created, what is electronic evidence, how electronic artifacts are forensically gathered, and the ability to analyze data to produce comprehensive reports and link analysis charts.

Our hypothesis is that Locard’s Exchange Principle does apply to cyber crimes involving computer networks such as identity theft, electronic bank fraud, or denial of service attacks, even if the perpetrator does not need to physically come in contact with the crime scene. Although the perpetrator may make virtual contact with the crime scene through the use of a proxy machine, we believe he will still “leave a trace” and digital evidence will exist. This presentation will explore with audience input “where in the cloud is digital evidence found” and new ways it can lead to attribution. It will explore what new standards and techniques are needed to find these digital traces. Read ahead information can be found at here.

Ken Zatyko was previously the Director of the Department of Defense Computer Forensics Laboratory where he led the largest accredited, internationally recognized, leading-edge computer forensics laboratory located in Maryland. For several months, Mr. Zatyko has been working with NIST on a working group to further standards and technology to solve cloud forensics challenges. Mr. Zatyko is currently the Vice President of Maryland Operations with Assured Information Security.

Host: Dr. Alan T. Sherman,

Computer Science and Electrical Engineering
University of Maryland, Baltimore County

Ph.D. Dissertation Defense

Multiple Objective Task Scheduling
for Scalable High Performance Computing

Tyler A. Simon

12:30-2:30 Friday, 8 November 2013, ITE 325b

Individual processor frequencies have reached an upper physical and practical limit. Processor designs now involve adding multiple processing elements to a single chip to enable increased performance. It is expected that all future processor designs will have multiple cores with a focus on reducing frequencies and adding more individual processing elements (cores) while having to balance power consumption, reliability and maintain high performance.

Due to the increased complexity as well as increased heterogeneity of parallel architectures, petascale and future exascale systems, with the number of processors on the order of 10^8-10^9, must incorporate more intelligent software tools that help manage parallel execution for the user. I demonstrate that by managing the parallel execution environment at runtime, we can satisfy performance tradeoffs for a particular application or application domain for a set of common HPC architectures. It is expected that future exascale computing systems will have to execute programs on many individual and potentially low powered processing elements. These processors need to be fed data efficiently and reliably through the duration of a parallel computation.

In this thesis I provide a performance analysis of two common graph algorithms for finding a minimum spanning tree and evaluate the multicore performance of a common high performance computing (HPC) benchmark on multicore processors. I also develop a novel autonomic execution model and adaptive runtime system (ARRIA) Adaptive Runtime Resource for Intensive Applications. ARRIA is designed with the intent of improving application programmability, scalability and performance while freeing the programmer from explicit message passing and thread management. Experiments are conducted that evaluate ARRIA’s capabilities on data intensive applications, those where the majority of execution time is spent reading and writing either to local or remote memory locations. In my approach, I focus on developing task schedules that satisfy multiple objectives for clusters of compute nodes during runtime. This approach is novel in that it can control application performance and satisfy constraints that are solved using multi objective optimization techniques as the program runs. The development and implementation of the ARRIA runtime system and subsequent optimization criteria likely provide reasonable models for the exascale computing era.

The results of this dissertation demonstrate, experimentally, that for high performance computing systems, a dynamic, task based, parallel programming environment and scheduler can provide lower total workload runtimes and high utilization compared with commonly used static scheduling policies.

The UMBC ACM student chapter will host the second meetup of Baltimore Code Craftsmanship user group at 6:45pm on Thursday, October 24 in the ITE building on the UMBC Campus.

This meetup is for the students and software developers in the Baltimore area that care about the quality of their work and want to practice and improve their programming skills, share what they know and learn new things from others.

Members of the UMBC community can RSVP for the event by completing this form. More details can be found on the form or below. Participants outside of the UMBC community should register on the meetup site. Registration for this meetup is available for a limited number of participants.

This is a hands on coding user group with no presentations. Each meeting will be a dojo where we will go through a challenging software craftsmanship exercise that focuses on clean code, test-driven development, design patterns, and refactoring. We will pair up and practice on a kata in order to learn and apply the values, principles, and disciplines of software craftsmanship. Come with your laptop equipped with your favorite programming and automated unit testing environment. If you don’t have a laptop, come anyway, we will need only one laptop for every two people. Be prepared to pair up, learn, share and have fun!

The event is open to all UMBC students, however programming ability is required. Interested faculty members can join in too! This can also be a good opportunity to network with professionals from various companies and get yourself noticed for any job opportunities that exist.

If you have questions or suggestions, send email to Primal Pappachan () or Vladimir Korolev ().

The UMBC Council of Computing Majors (CCM) will meet from Noon to 1:00pm on Monday, October 21 in room 101 in the Administration building. CCM is a student organization representing undergraduate computer science and computer engineering majors and anyone else with an interest in computing. Everyone is welcome.

At next week’s meeting, Austin Murdock will give an overview of Raspberry Pi and Arduino technology and lead a discussion about possible CCM group projects involving them.

The Raspberry Pi is an inexpensive, credit-card-sized single-board computer that runs Linux and Arduino is a single-board microcontroller used to make using electronics in multidisciplinary projects more accessible. Their low cost, size and power requirements make them ideal for building novel embedded applications in a familiar Linux environment using open source software.

The results of the UMBC Homecoming Spirit Office Decorating Contest were revealed last night at the Faculty/Staff Social. More than 30 offices participated this year and the CSEE Department’s entry took first place in the Judge’s Choice category, which was based on site visits by a panel of judges from the UMBC alumni association, and third place in the People’s Choice category, which was determined by an online vote. The first place showing in the Judges Choice competition comes with a trophy and ten tickets to the Homecoming community picnic.

Congratulations to the CSEE staff for the great job of decorating the ITE325 suite for UMBC’s Homecoming. See pictures of their Maryland Beach Picnic themed display and stop by ITE325 during the rest of this week to experience the full multimedia effect and let the Department’s staff know what you think of their effort.

UMBC’s CyberDawgs placed second at the 2013 Maryland Cyber Challenge (MDC3) finals that were held in Baltimore on October 9 as part of the CyberMaryland 2013 conference.

CyberDawgs-1 (pictured above) included William Thomas, Anthony Sasadeusz, Kenneth Johansen Jorge Teixeira, AJ Hallameyer, and Fahad Alduraibi.

During the finals, teams competed in a Control-the-Flag event where they attempted to gain (and maintain) control of other systems on the network – which were also being targeted by other teams seeking to gain and ‘own’ them as well.

All of UMBC’s three CyberDawg teams competed vigorously throughout the MDC3 season and maintained a very tight scoring situation for the college division. However, by taking second place in today’s finals, each member of CyberDawgs-1 will receive a $2,000 cash prize provided by the National Security Agency.

CSEE’s Dr. Richard Forno and Dr. Charles Nicholas are faculty advisors to the team.  UMBC is a co-founder of the Maryland Cyber Challenge.

If you are a computer science or computing engineering major, you should consider joining the Council of Computing Majors (CCM). This is a student organization that represents undergraduate CMSC and CMPE majors. Austin Murdock and Amanda Pyryt have revived the CCM and will lead the initial meeting for the year this coming Monday, October 7th from 12-1pm in Admin 101.

If you are interested in speaking with other Computer Science and Computer Engineering majors at UMBC, or interested in learning more about computing please feel free to attend! This year we are also introducing a system to set up study groups for your major.

The initial meeting will include a presentation from a Windows representative about Windows 8.1. This is the first major major update for the Windows 8 operating system and is expected to be released later this month.

UMBC Center for Information Security and Assurance

Quantum Computing and Cybersecurity

John Seymour

Noon-1:00 Friday, 4 October 2013
Cyber Defense Lab, room 228 ITE, UMBC

This talk will be a brief introduction to the topic of quantum computing for the computer scientist interested in cybersecurity. It will begin with a light summary of the fundamental quantum algorithms and move to discuss the recent advances in quantum computing, including the D-Wave quantum optimizer, University of Bristol’s new quantum chip, quantum programming languages, and more. Finally, it will introduce some current research questions and projects residing in the intersection of quantum computing and cybersecurity.

John Seymour is a Ph.D. student in the UMBC computer science graduate program. As a UMBC undergraduate, he was a triple major — Computer Science, — Mathematics and Philosophy. He is currently working on three research projects: evaluation of a detection protocol for Man-in-the-Middle attacks, a web-based game for teaching students basic concepts of internet security, and integration of social media with internet voting to facilitate collaborative decision making.

Host: Dr. Alan T. Sherman,