Computer Scientists at UMass Amherst Collaborate On Research Funded By $11.3 Million In Federal Grants

AMHERST, Mass. - Faculty in the department of computer science at the University of Massachusetts Amherst will collaborate on more than $11 million in grants from the National Science Foundation (NSF) to support four collaborative research projects focusing on marine science, data mining, computer networking, and network measurement infrastructure.

The funding was awarded through the highly competitive Information Technology Research (ITR) Program. More than 1,400 proposals were submitted for ITR grants, and only approximately 10 percent of the projects were funded.

Computer science faculty are also teaming with the electrical and computer engineering department in the University’s recently announced $40-million Engineering Research Center. Collaborative Adaptive Sensing of the Atmosphere (CASA) is one of only four new centers created this year. Funded in part by the NSF, CASA is expected to increase the warning time for tornadoes, flash floods, and other severe weather disturbances with far greater accuracy than existing systems.

“Our success with both the ITR and the ERC awards is a clear demonstration of the quality of our faculty and the computer science program at UMass Amherst,” says computer science chair Bruce Croft.

A description of each of the four NSF ITR projects follows.

* The UMass Amherst Computer Vision Lab is partnering with Bigelow Laboratory, a leading ocean science lab located in Boothbay Habor, Maine, on a five-year, $3.2-million grant for collaborative research in marine science. The grant is under the direction of professors Edward Riseman, Allen Hanson, and Paul Utgoff and Howard Schultz, senior research scientist.

According to Riseman, because the Earth’s oceans are under increasing threats of human activity, there is increasing need to effectively monitor suspended particles, both living organisms (plankton) and non-living detrital particles, using remote, in-the-water, continuous ocean observing systems and remotely operated or autonomous instruments.

The current project focuses on the development of environmental monitoring tools for marine scientists that, among other things, would allow early detection of harmful algal blooms along the coast, says Riseman. It addresses the needs of marine scientists across a multitude of problems for automated and interactive tools to process and interpret the huge numbers of images requiring expert interpretation. Riseman says the UMass/Bigelow team will produce common software tools that can effectively address the specific challenges unique to the varying imaging instruments and classification needs of marine science applications. The researchers will combine the best aspects of knowledge used by human experts, computer vision techniques for extracting features of particles, including color, texture, shape and size, and machine learning for classification of complex inherently ambiguous particles.

* Professors Andrew McCallum and David Jensen have teamed up on a five-year,$2.6-million project titled “Unified Graphical Models of Information Extraction and Data Mining with Application to Social Network Analysis.” According to McCallum and Jensen, this project aims to improve the ability to data mine information previously locked in unstructured natural language text. Their research focuses on developing novel statistical models for information extraction and data mining that have such tight integration that the boundaries between them disappear, resulting in a powerful unified framework for extraction and mining. The new algorithms will be applied to the creation of two large-scale databases with useful, publicly available Web sites including

-- detailed information on scientific research
-- news articles and Web sites about public government initiatives.

Mining these databases will enable insight into government efficiency and the flow of scientific ideas, say the researchers.

* Computer science professor Micah Adler will collaborate on a five-year, $2.8-million project with researchers at Columbia University and Polytechnic University of New York to combine expertise in computer networking, the Internet, and distributed systems along with modeling, stochastic networks, optimization and distributed algorithms. Titled, “Peer-to-Peer (P2P) Networking Theory,” Adler expects the project will revolutionize the Internet and computing over the next 10 years. Adler says his research is motivated by two fundamental challenges in studying P2P networks:

(i) P2P systems are enormously complex and create a networking environment that is significantly different from the traditional client/server model, and

(ii) it is difficult to evaluate new design proposals through online experimentation and simulation. Thus, there is a significant need for theoretical evaluation, design and analysis of P2P systems.

* Computer science professors Jim Kurose, Brian Levine, Prashant Shenoy, and Don Towsley are teaming with UMass Amherst engineering professors Lixin Gao, Weibo Gong, and Tilman Wolf on the five-year, $2.7-million project “Hyperion - next generation measurement infrastructure and application use.” According to Kurose, this fundamental, cross-disciplinary research project is aimed at the design, development, and application-use of a next generation of distributed, high-performance passive network measurement infrastructure. The project will be divided into four efforts:

(i) the design, development, and prototype implementation of a new network measurement node architecture based on the use of next-generation network processor (NP) chips; (ii) the use of multiple Hyperion nodes to enable a much richer set of network-management and traffic-profiling capabilities;(iii) the exploration of a number of application-level uses of Hyperion nodes, including fault-detection, overlay support, and traffic characterization; and (iv) working with network providers to understand network-management needs, and to demonstrate Hyperion capabilities.

The research team will involve collaborations with the UMass Amherst Office of Information Technologies (OIT), the staff of the Massachusetts Information Turnpike Initiative (MITI), and Sprint. Says Kurose, “We’re excited to have won this award. The five-year timeline, and the collaborations between departments and with network providers will give us a great opportunity to make fundamental advances here.”

Kurose, along with computer science professors Victor Lesser, Prashant Shenoy, and Don Towsley are also leading the research efforts in computational and networking aspects for the $40-million NSF ERC project.