• Undergraduate Computer Research (UnCoRe) for Women at Oakland University
• Fish Hook Immersive Educational Environment
• K-12 Educational Software for Palm-sized Computers
• Learning Online Network with CAPA (Computer-Assisted Personalized Approach
• Recruitment and Retention of Faculty in Computer Science and Engineering
  

Undergraduate Computer Research (UnCoRe) for Women at Oakland University

Ishwar K. Sethi and Fatma MiliOakland University, Rochester, MI

Examples of images classified by the REU Site research team at Oakland University

Aiyesha Ma, a senior computer science major, was a summer intern at the CISE-supported Research Experiences for Undergraduates (REU) Site at Oakland University (OU) in 2002. Along with research partner Rishi Mukhopadhyay, a freshman from Cornell University, Ma worked in the Intelligent Information Engineering Laboratory under Dr. Ishwar Sethi. They developed image classification tools based on perceptual vector quantization. Her innovation was the use of the Hausdorff measure of distance between sets of points on the images. Their results increased the perceptual robustness of “codeword blocks”—that function like words in text searches of typical search engines—in order to facilitate the indexing of collections of images. Their tools were able to differentiate between images of man-made objects with many straight lines (e.g., bridges) and images of natural objects without straight lines (e.g., cats). The results of their summer work were presented at the 2003 International Conference on Information and Knowledge Engineering. Ma is now a computer science master’s student at OU, and plans to earn a PhD. She was recently awarded a Michigan Space Grant Consortium fellowship based on her proposal to develop methods to automatically index satellite images.

Fish Hook Immersive Educational
Environment

Brian M. Slator in collaboration with Jeffrey Clark and the NDSU Digital Archive Network for Anthropology (DANA)

The Fish Hook Immersive Educational Environment team at North Dakota State University(NDSU) is developing a Virtual Archeology education simulation.Their work includes working with staff at the North Dakota State Historical Society Museum who are providing records and artifacts, and the Three Affiliated Tribes (Mandan, Hidatsa, and Arikara). The group of professors and students are developing a virtual reconstruction of an historical native American Indian village site -- first as it existed in 1954 when it was lost to flooding, and eventually as it existed in 1851 when it was inhabited by the Three Affiliated Tribes. This reconstruction will serve as the basis for an educational environment in the scientific aspects of archeology, with the digital library resources providing images and 3D models in an Internet-accessible archive.

Work on this project so far has included 3D laser scanning of several dozen artifacts recovered from the site, 3D reconstruction of terrain and earth lodges based on the archeological record, development of a 'to the millimeter' text-based simulation of the surrounding environment, initial modeling of flora in the environment, and the compilation of an extensive bibliography of materials and records, see the project web site at: http://fishhook.cs.ndsu.nodak.edu/

K-12 Educational Software for Palm-sized Computers

Elliot Soloway et al.

As we move into the 21st century where knowledge-work and global connectivity is becoming paramount, there is an even greater urgency 'to leave no child behind.' Thecomputing industry is producing low-cost, palm-sized computers that can be placed in the hands of each and every K-12 child. Hardware ubiquity is a necessary but not a sufficient condition for impacting K-12, the new challenge is putting appropriate educational software on those palm-sized computers.

An ITR-sponsored project headed by Elliot Soloway at the University of Michigan's Center for Highly-Interactive Computing in Education is developing a 'learner-centered design' framework that guides the creation of the next generation of educationalsoftware that will run on wireless-networked, palm-sized devices. For example, 7th grade students, with such connected palm-sized computers nested in their hands all day in school, harvest key Internet-based resources, and work collaboratively in coming to understand science, mathematics, social studies, and language arts. Currently thousands of K-12 students around the U.S. are using these palm-computer educational applications, e.g., Cooties is a socio-kinesthetic simulation program that enables children to better understand how germs and viruses are spread; picoMap is concept mapping program that enables children to create visual outlines; Sketchy is a animation program that enables children to depict dynamic science experiments; Go 'N Tell uses a camera attached to the palm-computer to enable children on a field trip, say, to capture images and then annotate them.

Learning Online Network with CAPA (Computer-Assisted Personalized Approach)

Gerd Kortemeyer et al.

Michigan State University's Learning Online Network with CAPA (Computer-Assisted Personalized Approach) is building an open platform for the development and distribution of online content. The network, dubbed LON-CAPA, would make online content freely and openly available to any instructor in the sciences or social sciences. Gerd Kortemeyer, leader of the LON-CAPA team, describes it as a digital library with a built in instructional management system. Currently, it includes material for courses in physics, calculus, chemistry, biology, food science, and psychology. The platform offers automatic checking of homework problems, with helpful feedback available to those who come up with incorrect answers.

The LON-CAPA project last year gained four Michigan high school science teachers as collaborators in an NSF-sponsored 'Research Experience for Teachers' project. Over the summer the teachers developed curricular materials and online exercises at Michigan State University (MSU), which were then deployed during the school year using networked servers installed in their high schools. Several of these developed resources were at a level such that they were also used for courses at MSU, while some of MSU's own curricular material was used at the high schools.

Recruitment and Retention of Faculty in Computer Science and Engineering

Bill Aspray and Jack Stankovic

Investigating and documenting the environment, processes, and prospects for the retention and recruitment of faculty in computer science and engineering was the goal of this project headed by Bill Aspray and Jack Stankovic of the Computing Research Association.

To accomplish this, the project examined and detailed the process of faculty recruitingfrom the perspective of new PhDs, established faculty,department chairs, and faculty search committees. To add hard data to previously anecdotal accounts, the team constructed and administered six surveys, each targeting different actors in the recruiting and retention process. This created a fuller picture of the recruiting environment than existed before.

This study provided essential information on the recruiting and retention environments, information that will allow universities and undergraduate colleges to make better-educated decisions about how to attract and keep high-quality faculty members.

ITR/AP(DEB): Collaborative Research Computing Optimal Phylogenetic Trees under Genome Rearrangement Metrics

Robert Jansen (UT), Bernard Moret (UNM), David Bader, Tandy Warnow

The goals of this project headed by researchers at University of Texas at Austin and University of New Mexico is to develop, implement, assess, and refine high-performance algorithms for phylogeny reconstruction from gene-order and gene-content data.

New distance estimators for genomic distances based on gene orderings, along with novel encodings of gene orderings into sequences, have led to the development of an array of entirely new techniques to reconstruct accurate phylogenies within reasonable time.

GRAPPA, a software package for phylogeny reconstruction from gene-order data, has demonstrated a one-billion-fold speed-up over prior packages for the same task, with much more accurate results.