|
|
|
|
|
|
|
|
The department has broad expertise in four flagship research areas:
Seminar and Colloquium Information
Upcoming Conferences at UTEP:
SIGDOC 2007
... and many more, many listed on individual
faculty pages.
An Invitation: The faculty and students involved in each project
typically have weekly meetings, which have not only project-related
activities but also presentations and discussions of papers and
systems. If you're interested in participating, contact a faculty
member for details.
There are also many smaller research projects.
Several have arisen from interdisciplinary collaboration;
others reflect student initiative, such as work in universal
access, in computer science education, and in computer-mediated communication.
These and other projects are supported by the NSF (including a
large MII grant), DARPA, DoE, USCGS, ARL, NASA,
and IBM, among others.
Sample Projects
In extreme scale
systems, which consist of hundreds of thousands of nodes, conventional
operating systems become a performance bottleneck. Dynamically
changing resource usage patterns demand OS dynamic adaptation.
We are exploring various adaptations, for example in I/O scheduling,
and ways to develop or evolve heuristics to trigger adaptation (DAiSES Homepage).
Pat Teller, Barton Miller (University of Wisconsin).
Peer-to-peer systems need to develop structure based on the observed
behavior of the nodes they encounter. We are exploring applications
of game theory and constraint processing
to solve problems of data placement and trust
determination.
Eric Freudenthal,
Francois Modave,
Martine Ceberio.
We are working to improve the usability of spoken dialog systems, to enable users to
access information over the telephone more easily and quickly, especially
by improving the turn-taking to make system responses swifter and more appropriately timed
(Interactive Systems Group,
Responsive Systems Project).
David G. Novick,
Nigel Ward ,
Thamar Solorio,
Olac Fuentes,
Jon Amastae (Linguistics), Bill Lucker (Psychology).
One way to make software more reliable is to create detailed
specifications and check that the implementation in fact meets these
specifications. We are developing JML, an extension to Java, which
allows automatically checkable, executable specifications to be
embedded in the code (UTJML Home Page).
Yoonsik Cheon, Gary Leavens (Iowa State University).
The Cassini spacecraft
lifted off in 1997 and began a four-year mission orbiting Saturn in
2004. During this time, three software products primarily developed
and maintained at UTEP have been released to the rings science
community. These include the Footprint Viewer, a tool
developed to assist in the detailed construction and visualization of
observation sequences; this was responsible for identifying
previously undiscovered errors in observation sequences. Currently
we are extending the CASVU data analysis tool, which was used in
2004 for the Cassini Orbital Insertion Hazard Analysis where data
collected during the approach phase of the mission was used to assess the
hazard to the spacecraft as it crossed the ring plane.
(Cassini-Hugyens Mission,
Planetary Rings Node,
Cassini Orbital Insertion).
Steve Roach.
Predicting the structure of proteins in vivo is a challenge because of
the vast complexity of the problem. We are exploring algorithms based
on flexible constraints and various machine learning methods
which have the potential to drastically reduce
the amount of computation required.
Francois Modave, Martine Ceberio,
Olac Fuentes.
Reasoning under uncertainty is a pervasive problem in science and
engineering. We are applying and extending interval computation
techniques to problems in geoinformatics, image processing,
hardware timing analysis, finite element methods, reliability
assessments, and medical treatment, among other areas
(Interval Computation Web Site).
Vladik Kreinovich,
Martine Ceberio.
Virus scanners don't have enough choices:
a program is either banned or fully trusted, so systems remain
vulnerable to "malware" not yet known to be dangerous. We are
exploring the extension of non-discretionary access control techniques
to make them suitable for containing damage to un-administered (home)
systems.
Eric Freudenthal,
Luc
Longpre.
Divergences between formal specifications and the code written to
implement them are a major source of errors in complex systems. The
use of transformations is one way to automatically derive the
implementation from a specification, thus avoiding errors. However,
transformations can be difficult to write. Our work focuses on ways to
demonstrate that a set of transformation rules do indeed preserve the
semantics of a specification. So far we have verified some of the
software used in the implementation of the Sandia Secure
Processor. This software was implemented using transformations written
in HATS, originally developed by Victor Winter. We verified the
transformations using the automated theorem prover ACL2. We are now
planning to extend this proof system to handle more general
transformation rules.
(HATS,
ACL2).
Steve Roach.
We are pioneering the new face of cyber-infrastructure by providing
access to a diverse collection of large-scale resources and software
tools to support various kinds of geological analysis and research
(Geon Project Homepage).
Ann Quiroz Gates,
Randy Keller (Geology).
We are developing a unified and comprehensive approach to the verification and
validation (V&V) of software systems, integrating V&V of
both functional and non-functional properties. This project is aiming to
engender a major enhancement in
software quality, by enabling an approach where
properties can be verified by static analysis, testing, or model
checking or be compiled to runtime monitors as appropriate or
required
(Milaap Homepage).
Yoonsik Cheon,
Pat Teller,
James Browne (UT-Austin),
Calvin Lin (UT-Austin).
We are pioneering the new face of cyber-infrastructure to foster
interdisciplinary research, education, and professional collaborations
in science through access to a diverse collection of large-scale
resources and software tools. Our research focuses on service-oriented
architecture (SOA) technologies that support the publication,
discovery, and usage of resources over a grid, with an emphasis on
formal methods for checking correctness of data and web services,
provenance (metadata about information origin) languages, tools and
methods in support of trust and accuracy for results from
cyber-infrastructure-based applications and services. The provenance
infrastructure is based on Inference
Web Web technologies. We are collaborating and applying this work
to a number of projects including the Geoscience Network (GEON Project Homepage) and the Circumarctic Environmental Observatories Network
(www.ceoninfo.org).
Software Engineering Affinity Laboratory (SEAL).
Ann Quiroz Gates,
Vladik Kreinovich,
Paulo Pinheiro da Silva,
Steve Roach.
Developing rational policies for the
release of information from databases requires good models of privacy.
We are developing methods for ascribing privacy values to individual
pieces of data and for computing the value of a collection of
pieces of data.
Luc
Longpre.
Support
| CS Home | cswebmaster@utep.edu |