Large Scale Distributed Information Systems
Director: Prof. Amit Sheth
Dept. of Computer Science, University of Georgia
415 Graduate Studies Research CenterAthens, GA
Tel. (706) 542-2310 FAX: (706) 542-2966
Statement and Lab Profile
Most recent review of LSDIS Lab’s
research appears in:
A. Sheth, J. Miller, K. Kochut, and B. Arpinar, Research in Multi-Organizational Processes and Semantic Information
Brokering at the LSDIS Lab, SIGMOD Record, 30(4),
Available in: gzipped
Postscript, PDF , Word,
Two of the
most active research projects are:
Semantic Association Discovery
(which builds upon research in the Enabling Infocosm
Theme), and METEOR-S: Semantic
Web Services and Processes, Applying Semantics in Annotation, Quality of
Service, Discovery, Composition, Execution (which builds upon the research
in the Enterprise Integration theme).
Both projects belong to the emerging area of Semantic Web.
following overview predates the above.
Scale Distributed Information Systems Lab (LSDIS) at the University of Georgia was established in
the Fall of 1994 to perform research and technology development on two
complementary themes in distributed information systems: Enabling Infocosm
and Enterprise Integration.
Enabling Infocosm Theme. The
merging of computers and communications with significant advances in
networking infrastructure has made millions of information sources with
wide varieties of information accessible. This capability gives us a vision
of an information rich society "infocosm", where we expect to have any
information any where we want in (m)any form(s) for effective decision
making and knowledge-centric activities, improved productivity, and fun.
The focus of the work in this area is on creating an infrastructure for managing
heterogeneous information and information brokering. Our current work in
this theme is the InfoQuilt project, with the associated or past
projects ADEPT, InfoHarness/VisualHarness, and VideoAnywhere.
InfoQuilt. This project intends to develop a comprehensive
information brokering environment for enabling the Infocosm.
It focuses more on the logical and semantic issues though support for a
large variety of metadata (content independent, content descriptive,
content dependent), context, user models and ontologies for an environment
consisting of heterogeneous data sources. Current work includes
logical metadata correlation of information through Metadata Reference
Links (MREFs, as extension to HREFs),
media-independent search of heterogeneous media (structured data, text,
image, video) repositories, context
representation, and support for multiple (possibly preexisting)
ontologies. The InfoQuilt system also
intends to prototype a second generation World Wide Web architecture that
would be fully distributed, highly scaleable and extensible utilizing agent
technology where the information is managed and manipulated at logical and
semantic level (rather than physical hypertext link).
ADEPT [Alexandria Digital
Earth Prototype Modeling System]. The ADEPT
project is a comprehensive follow-on to the previously funded Alexandria Digital Library Project (ADL).
ADEPT aims to use the
Digital Earth metaphor for organizing, using, and presenting information at
all levels of spatial and temporal resolution. We call the digital
environments based on the Earth metaphor Iscapes
(Information Landscapes). The research and development activities focus
on: (a) creating geospatial information and meta-information collections;
(b) building operational services for: discovering heterogeneous,
distributed collections; organizing these resources into Iscapes tailored for specific applications; and
collaborative use and visualization of Iscapes;
(c) applying and evaluating ADEPT services
in learning environments; and developing scalable, efficient, and secure
systems. This project is supported by the National Science Foundation under
Grant No. IS IRI-9411330.
supports search and retrieval of heterogeneous information in
intranet/Internet environments. The basic InfoHarness
system, and the corresponding commercial product Adapt/X Harness from Bellcore, provide access to heterogeneous textual and
semi-structured data without restructuring, reformatting and relocating the
data. It also supports logical restructuring of the information
space, support for multiple third-party indices and many other features.
Research in the joint Bellcore-LSDIS project
funded by the Massive Digital Data System initiative addressed several new
research issues, including: keyword- as well attribute-based access to
data, access to multiple autonomous repositories with independent and
heterogeneous indices for scaleable search with associated intelligent
merging of results, access to the same repository with multiple indices to
improve quality of result, and support for remote severs using CORBA.
VisualHarness system is the InfoHarness
system enhanced with Zebra component, which provides customizable and
extensible search of federated image repositories, in addition to text and
semi-structured data, and structured (relational) databases. Through the
comprehensive use of metadata of various types, it supports keyword,
attribute-based and content-based search over images.
VideoAnywhere. This industry-funded project researches
the issues related to management of video in embeded
(e.g., cable set top) system running a Java Virtual Machine. It fits
into the exciting and fast evolving interactive TV and video application
market-place. It involves research in understanding metadata and consumer
profile, caching, crawling (using agent technology), etc.
Integration Theme. The interoperability of existing and new information
systems within and across enterprises has important implications on its
competitiveness. Three important components of a comprehensive approach to
enterprise integration are: heterogeneous information integration (also see
the next theme), coordination through process/workflow management, and
collaboration. Projects supporting this theme are as follows (two of the
projects, CaTCH and CareWeb,
also support the theme of telemedicine).
METEOR [Managing End-To-End OpeRations]. The METEOR system focuses on R&D
of innovative Multiparadigm Transactional
Workflow Management technology. Workflow management techniques and systems
developed in this project support coordination of user and automated tasks
in real-world multi-enterprise heterogeneous computing environments over
Web/Java and CORBA infrastructures. Technical issues involve GUI toolkit
for workflow design/monitoring/simulation, automated code
generation for distributed workflow application from graphical design, scheduling,
EDI, multidatabase access, error/failure
handling and recovery (using transactional concepts/techniques) as
appropriate, and security. This project is funded through the NIST ATP
initiative in Information Infrastructure for Healthcare and involves over
17 companies and R&D institutions. We also assist our main
partner (Connecticut Hospitals Assoc.) in
developing real-world applications. METEOR software has been tested by our industry partners and
can be sublicensed from a small company for commercial applications. METEOR designer has been made
available free of charge to instructors conducting advanced graduate courses
in this area, and METEOR technology
has been licensed for commercialization by a private company.
projects supporting this theme follows:
CaTCH [Collaborative Teleconsulting
This project involves integrating multimedia patient data on Intranet,
medical reference data on Internet, LAN/POTS/ISDN-based Video+Data Conferencing, and WWW/Java programming to set up
remote environment and context sensitive collaboration. Medical College of Georgia is our primary
application partner for this project. One version of the system is
being field trialed.
CAPA [Course Approval Process Automation] is an operational,
production workflow system developed and supported by the LSDIS. It
supports a 30+ step workflow spread over entire UGA campus with potential
user-base of 130+ departments and 3500+ faculty and staff members. It
has already been used to support approval of 6000 new courses in a
CareWeb. It is an advanced
Web-based system developed in the Georgia's Family Connections
program for the Medical College of Georgia to support distance health
education and collaboration among patients and healthcare workers.
Relevant Information on the LSDIS Lab
lab is directed by Professor Amit Sheth. The Lab's
activities involve active participation of Professors Krys
Kochut and John Miller. Lab’s
research funding of approx. $3 million over the last four years.
Students working in the lab frequently do internships in industry during
their study, participate in international conferences and are aggressively
pursued by potential employers. Through its collaborations with
industry partners, the lab is able to achieve significant technology
transfer. Boeing, MCC, LGERCA and Hewlett
Packard Labs have joined the LSDIS Lab as industry
sponsor/affiliates. The lab has extensive interactions with
healthcare industry, with recent interactions with manufacturing and consumer
been a member of Healthcare Open Systems Trials (HOST) , the Object Management Group (OMG) and the Workflow Management
The LSDIS has been supported by significant product donations (exceeding $400,000
in value) from Informix, Iona Technologies, Virage, I-Kinetics,
PictureTel, Persistence Software, Lucent Technologies, etc.
LSDIS's computing resources support distributed
and collaborative computing, Intra- and Internet environments, wired and
wireless LAN communications, and heterogeneous media data and multimedia
document creation and management. It has approx. 40 computers
including seven servers, network storage, wireless and wired LAN, and a
large variety of latest relevant software products for developing and
supporting distributed, Internet-based and multimedia information
you to see more up-to-date details (including demonstrations, publications,
etc.) at http://lsdis.cs.uga.edu