M.S. Thesis Defense Announcement

Feature Extraction using a Hierarchical Growing Neural Gas

Roger Guseman

12:00pm 25 April 2011, ITE 210

Unsupervised, data-driven, automatic feature extraction from image data is an interesting and difficult problem. High dimensional data, such as images, often contain less information than they do data. For an agent to better reason about this data, finding the "interesting" features in the data is helpful. A current technique, known as the Growing Neural Gas (GNG), is a neural network approach to feature extraction. There are, however, adaptations that can be made to the Growing Neural Gas in order to increase its performance.

Contributions of this work include development of a new neural network algorithm extending the Growing Neural Gas framework, known as the Hierarchical Growing Neural Gas (HGNG), identification of how the parameters of the HGNG affect feature extraction performance, and theoretical and empirical comparisons of performance between the normal GNG and the HGNG neural networks.

Thesis Committee:

• Dr. Tim Oates (chair)
• Dr. Tim Finin
• Dr. Marie desJardins

Master's Thesis Defense Announcement

Problem Selection of Program Tracing Tasks in an Intelligent
Tutoring System and Visual Programming Environment

David Walser

2:00pm Thursday, 28 April 2011, ITE 325b

Intelligent tutoring systems (ITSs) have been shown to be an effective supplementary teaching tool or aid for many domains. Applying ITSs in open-ended domains such as computer programming is especially challenging, most notably when trying to assist with the process of programming itself. Existing ITSs for programming focus on a very limited set of problems and concepts and are only useful early in an introductory CS course and a few limited places afterward. Visual programming environments are another tool that have been used in introductory CS courses to help students learn basic concepts. The key idea behind my work is the recognition of the importance of students' ability to read, understand, and trace code in order to write programs successfully. A broader goal of my work is to show that an ITS based on a visual programming environment can be used to support students throughout an entire introductory CS course, without being severely constrained and limited to a small number of concepts and to low-level, simple tasks. In my system, called RUR-ITS, students are given a program and are asked to predict the robot's behavior when running this program in a given environment. RUR-ITS allows each problem to be assigned a difficulty level and multiple concepts that it involves within the conceptual model. RUR-ITS can then use a problem selection algorithm to choose a problem that is most able to help the student master the concepts that they have not yet mastered.

Thesis Committee:

• Dr. Marie desJardins, Chair
• Dr. Tim Finin
• Dr. Tim Oates

   

USA Today has an article on the increased recruiting by technology companies, Tech jobs boom like it's 1999. Potential employees are being lured with good salaries, bonuses, and perks, and some predict that almost 150,000 technology jobs will be added in 2011.

"Nearly 150,000 tech jobs are expected to be added this year, says Sophia Koropeckyj, an economist at Moody’s Analytics. In February, there were about 6.1 million tech jobs in the U.S., up 2.4% from a year ago. Yet there’s a lot to make up for: From the second quarter of 2008 through the first quarter of 2010, during the economic swoon, 308,000 tech jobs were lost."

Companies such as Google, Facebook, and Microsoft are recruiting students to try to hire them before they graduate, says former Google software engineer Gayle Laakmann McDowell. Some economists see the boom in tech hires as a good indication that the overall U.S. economy is on the rise, since technologies such as wireless, computing, and energy have a big impact on other sectors.

"A surge in tech hires in California could portend an upturn for the overall U.S. economy, says Jesse Harriott, chief knowledge officer at online job site Monster.com. That’s in part because what’s good for tech will touch related industries. Areas such as wireless, computing and energy have an effect on “a substantial part of the economy, such as logistics, inventory management, enterprise coordination, even hybrid autos, says Shane Greenstein, a management professor at Northwestern University’s Kellogg School of Management. Entrepreneurs and venture capitalists foresee growth in wireless apps, online gaming and clean tech, he says. “It is not a classic tech boom, led by a big new opportunity like (Internet) browsing or Web 2.0, but a mixture of a few big and unrelated trends,” Greenstein says.

We've seen the increased interest in our students here in the Computer Science and Electrical Engineering department and it includes both undergraduates and graduates. The demand is up not just for science, engineering, mathematics and technology majors but also for majors in the liberal arts and other areas who have acquired a portfolio of math and technology skills.

Amond Uwadineke is studying how the presence of oil in the Niger Delta has influenced conflict within this important region in Nigeria, and how this natural resource has hindered the development of the political institutions of the Nigerian State.

See Amond's presentation in the UC Ballroom on Wednesday, April 27 from 12:30 to 3:00 p.m..

Spring 2011 Information Systems Distinguished Lecture

Working Together Apart

Dr. Judy Olson

Donald Bren Professor of Information and Computer Sciences
University of California, Irvine

12:00pm Friday, 29 April 2011, ITE Lecture Hall VII

Our research group has been investigating the factors that make long distance teamwork work. I will review that work and talk about the issues that remain, the factors that technology and social practices can't solve–cultural differences and different timezones. And, I will introduce our "theory made practical," our turning the theory into an online assessment tool.

Judith Olson is the Bren Professor of Information and Computer Sciences in the Informatics Department at the UC Irvine, with courtesy appointments in the School of Social Ecology and the Merage School of Business. She has researched teams whose members are not collocated for over 20 years, summaries of which are found in her most cited paper, “Distance Matters,” (Olson & Olson, 2000), and in her key theoretical contribution in the book Scientific Collaboration on the Internet (Olson, Zimerman, and Bos, Eds., 2008). Her current work focuses on ways to verify the theory's components while at the same time helping new scientific collaborations succeed. She has studied distributed teams both in the field and in the laboratory, the latter focusing on the communication hurdles distributed teams have and the consequent underutilization of remote team members skills and the reduction in trust. She is a Fellow of the Association for Computing Machinery and with her husband and colleague, Gary Olson, holds the Lifetime Achievement award from the Special Interest Group in Computer Human Interaction.

· directions · talks · events ·

The Maryland Cyber Challenge and Conference site is up and student teams can now register for the competition, with the first qualifying round early in September. It is a chance to demonstrate your ability to work in a team and your cybersecurity and problem solving skills.

MDC3 is a joint effort between SAIC, UMBC, DBED, TCM and NCSA to bring people together to promote Maryland's commitment to cybersecurity and STEM education. The competition includes three levels: high school, collegiate and professionals from industry/government, providing opportunities to network with cybersecurity professionals, researchers, and scholars.

There will be orientation sessions at the UMBC Technology Center (1450 South Rolling Rd., 21224) on May 2, May 18 and June 21 at 4:30pm for professionals and 6:00pm for students.

One of the most critical steps to integrating heterogeneous e-Business applications using different XML schemas is schema matching, which is known to be costly and error-prone. Many automatic schema matching approaches have been proposed, but the challenge is still daunting because of the complexity of schemas and immaturity of technologies in semantic representation, measuring, and reasoning. The dissertation focuses on three challenging problems in the schema matching. First, the existing approaches have often failed to sufficiently investigate and utilize semantic information imbedded in the hierarchical structure of the XML schemas. Secondly, due to synonyms and polysemies found in natural languages, the meaning of a data node in the schema cannot be determined solely by the words in its label. Thirdly, it is difficult to correctly identify the best set of matching pairs for all data nodes between two schemas. To overcome these problems, we propose new innovative approaches for XML schema matching, particularly applicable to XML schema integration and data transformation between heterogeneous e-Business systems. Our research supports two different tasks: integration task between two different component schemas; and transformation task between two business documents which confirm to different document schemas.

Dissertation Committee:

• Dr. Yun Peng, Chair
• Dr. Charles Nicholas
• Dr. Zary Segall
• Dr. Milton Halem
• Dr. Hyunbo Cho (POSTECH, Korea)
• Dr. Nenad Ivezic (NIST)