The research that we are proposing is based in part on previous
research that has been reported in the literature, and in part
on our own work.
We note that in
[41], several experimental systems have been presented and
their use as prototype digital libraries discussed. WAIS, the wide area
information servers [35], can potentially connect thousands
of users and handle millions of queries against many bibliographic and
text databases. AT& Bell Laboratories Library Network [29] offers the RightPages
Service. This service provides
the user with an ``electronic window'' which combines text and bitmap
images by scanning actual journal pages into the system and displaying
them on the screen at the user's workstation. The system also
automatically alerts users to information published in their fields of
interest.
With respect to the related research at UCSB, there are several threads of relevance. First, the Alexandria Project investigators associated with the UCSB MIL (Carver, Larsgaard) recently completed research on a conceptual design for a geo-referenced information network (GRIN) [58]. The purpose of the system was to improve the management of, and access to, geographically referenced information, such as maps, aerial photography, and atlases. The interface allows the user to navigate in the database using a master locator map and select processing functions with menus. The graphic nature of the interface allows the user to learn the system with a minimum of training and preparation. Other graphic interface designs have also been proposed. For example, [20] presents an interface design based on the map-overlay metaphor. Data cubes are used to represent the spatial and attribute data. Templates describe rules of how to render the data in a cube. By moving and stacking cubes and templates graphically, the user is able to render geographical information in a way similar to conventional map overlay.
Second, several Alexandria project investigators from the UCSB Computer Science Department (Agrawal, El Abbadi, Smith, Su) have been engaged in an NSF-supported project to develop a Modeling and Database System (MDBS). The main goal is to develop a system that supports the iterative model-building efforts of earth scientists who are involved in data-intensive and computationally-intensive investigations. In particular, MDBS incorporates support for the scientific process of iterative model development; the construction, maintenance, and access of scientific databases; and the concepts of scientific models and collaborative projects. To date, the research effort has resulted in a data model for modeling geographic information, a high-level modeling language, database support and storage mechanism for complex elements, and a proof-of-concept system for earth science research groups at both the University of Washington and UCSB.
Third, the success of the NSF-sponsored project has resulted in an enlargement of this group which now includes some other Alexandria Project investigators working in the area of parallel processing and visualization (Egecioglu, Ibarra, Wang). The expanded group recently won a major 3-year award from NASA that will permit them to design and construct a full-scale MDBS. The goal of the project is to solve major EOS-scale problems, such as routing of water and sediment down the entire Amazon basin, the intergration of GCM's and regional hydrological models, and the modeling of snowmelt and routing of runoff from regional snowpacks.
Fourth, the Alexandria Project investigators from the NCGIA (Beard, Buttenfield, Egenhoffer, Goodchild, Mark, Onsrud) have taken part in a large number of NCGIA Research Initiatives over the past five years issues that bear directly on nearly every issue that must be addressed in relation to digital library support for spatially-indexed information. Fifth, the Alexandria Project investigators from the Center for Remote Sensing and Environmental Optics (Davis, Dozier, Frew) are also centrally involved in the Sequoia 2000 project. This project may, in a real sense, be viewed as a forerunner of digital libraries for scientific investigators.
Finally, Alexandria Project investigators associated with the CIPR at UCSB (Manjunath, Mitra) are involved in research on image processing and analysis. The research in CIPR has a strong emphasis on efficient algorithms, techniques, and architectures for data compression, image and video processing, and multi-sensor estimation. Current research topics relevant to the Digital Library Initiative include various image processing, coding, and compression algorithms, such as non-linear algorithms for image enhancement, image zooming and printing, image registration, and sensor fusion.
Clearly the above projects are of direct relevance to the Digital Library Initiative. Issues such as interface design, parallel computing support, data modeling, and database support for spatially-indexed information are being addressed in these projects. Furthermore, the experience of the Alexandria Project investigators involved in the above projects is greatly strengthened by other investigators who bring important skills from relevant areas of computer science (Cappello, Dillon, Gonzalez, Kemmerer, Konheim, Singh, Yang). In summary, we believe that the research background of the Alexandria team as a whole provides an outstanding basis for the Alexandria Project.