My research interests include scientific database
systems, access methods, data clustering, and software architecture. This
research is related to three ongoing efforts in computational sciences that
will have significant impact on future scientific research: the development of data
storage systems, which must provide high-performance access to data regardless
of the storage media and the way it is interconnected; the development of
advanced data-mining techniques, which will serve as primary means of analyzing
scientific data; and the development of scientific data grids, which will
provide a generic platform for widely distributed extraction of complex
information from extremely large collections of scientific data.
The broad objective of this research is to advance
the science of database systems in a direction that will illuminate the
enormous problems of scale in the management of persistent data and lead to appropriate
solutions. Due to the emerging petabyte data stores, growing dimensionality of
data, and mounting complexity of DBMS development, these problems of scale
attract considerable interest. An appropriate solution to a problem of scale
must gracefully adapt to a wide range of the increasing magnitudes of the
problem. The issues that have attracted most of my attention include the
problems of data volume, data dimensionality, system load, and DBMS development
complexity.
Data volume. A broad infrastructure for scientific research must provide
efficient solutions for widely distributed extraction of complex information
from extremely large collections of scientific data. One of the objectives of
my research is to support this infrastructure by developing original techniques
for efficient storage and retrieval of massive scientific data. To learn more
click here.
Data dimensionality. Another objective of my research is to
systematically analyze and effectively attack the limitations of contemporary
retrieval and clustering techniques in spaces with many dimensions. In this
area, I have been designing new access methods and clustering techniques for
high-dimensional data. To learn more click here.
System load. Part of my work involves the design of new transformation
schemes for the deployment of advanced indexing techniques in transactional
DBMS environments. The goal is to enable a rapid deployment of
multi-dimensional access methods by reusing the existing indexing techniques of
transactional DBMSs and to develop efficient concurrency and recovery protocols
specialized for compact indexing structures. To learn more click here.
DBMS development complexity. A stumbling block for component DBMS
construction is a high potential for mismatch between the DBMS artifacts. In
this area, my focus has been on the definition of a formal model of software
architecture that can be used to reason about architectural mismatch and on the
development of a formal design methodology that can guide the designers of
component DBMS architecture in their attempts to uncover and prevent potential
mismatch. To learn more click here.
·
J. Lukaszuk and
R. Orlandic, “On Accessing Data in High-Dimensional Spaces: A
Comparative Study of Three Space Partitioning Strategies,” Journal of Systems and Software 73(1):147--157, 2004.
·
R. Orlandic and B. Yu, “Scalable
QSF-Trees: Retrieving Regional Objects in High-Dimensional Spaces,” Journal of
Database Management 15(3):45--59, 2004.
·
J. Lukaszuk and
R. Orlandic, “Efficient High-Dimensional Indexing by Superimposing
Space-Partitioning Schemes,” Intern'l
Database Engineering and Applications Symposium IDEAS'04, Coimbra, Portugal, 257--264, 2004.
·
R. Orlandic and
Y. Lai, “Clustering Technology of a Data Engine for Analytical Computing,” Proc. IEEE 4th Intern'l Conf. on
Intelligent Systems Design and Applications ISDA'04, Budapest, Hungary,
699--704, 2004.
·
R. Orlandic, “ITR:
Development of a Data Engine for Grid-Enabled Analytical Computing”,
NSF Information and Data Management Workshop IDM’2004, Cambridge, MA, 3
pages, 2004.
·
R. Orlandic, “Retrieval
and Clustering of High-Dimensional Scientific Data,” Proc. 7th Workshop on
Mining Scientific and Engineering Datasets, Lake Buena Vista, Florida, 1
page, 2004. (abstract of the invited keynote speech)
· R. Orlandic, “Effective Management of Hierarchical Storage Using Two Levels of Data Clustering,” Proc. 20th IEEE / 11th NASA Goddard Conference on Mass Storage Systems and Technologies MSST’2003, San Diego, CA, 270--279, 2003.
·
B. Yu, T. Bailey, R. Orlandic and J. Somavaram, “KDBKD-Tree:
A Compact KDB-Tree Structure for Indexing Multidimensional Data,” Proc. Intern'l Conf. on Information Technology:
Coding and Computing ITCC'2003,
Las Vegas, Nevada, 676--680, 2003.
· J. Lee, D. Grossman and R. Orlandic, “Adopting a Hierarchical Category Dimension into A Multidimensional Information Retrieval Engine,” Proc. Intern’l Conf. on Information and Knowledge Engineering IKE’2003, Las Vegas, Nevada, 7--10, 2003.
· J. Lee, D. Grossman and R. Orlandic, “An Evaluation of the Incorporation of a Semantic Network Into a Multidimensional Retrieval Engine,” Proc. 12th Intern'l Conf. on Information and Knowledge Management CIKM'03