|
|
Mobile Information Management (MIM)
Executive Summary
This research investigates systems that attempt to resolve the
scalability, heterogeneity, independence, and variability problems in
large-scale distributed information integration.
The approach constructs multiple layers of abstraction in the form of
autonomous components and focuses on flexible optimization that exploits
redundancy in information pathways.
The autonomous components considered exchange and manage data,
queries, query plans and distributed query processes in the form of
mobile information.
Such components, dubbed Mobile Information Managers, rely heavily on
optimization techniques that speed up query processing by using
specialized access structures.
An interesting kind of optimization chooses between distributed plans
that decide to perform computations in different sites.
These optimization depend on a cost model that takes into
consideration connectivity, i.e., performance and reliability in
networks.
Advances in networking and in computing technology are often changing
the parameters of the model.
By demonstrating the feasibility and joint impact of several kinds of
optimization needed in large-scale distributed information integration
applications, the project hopes to constribute to better software
engineering principles and ideas for such applications.
Several very active areas could make esential use of such
applications: electronic commerce, digital libraries, scientific
databases.
This broader potential enhances the significance and the timeliness of this work.
Project Members
Val Tannen Alin Deutsch Lucian Popa Arnaud Sahuguet
Publications
|