OVERVIEW OF PROGRESS IN LIBRARY DEVELOPMENT

We rate our progress in terms of both research and development as very good, and significantly beyond our expectations at the beginning of the Project.

In terms of the history of the development of our testbed system,

1. during the first six-months of the Alexandria Project, (10/01/94-02/28/95), we successfully completed our first testbed system, which was a stand-alone ``rapid prototype'' (RP) system;
2. during the next twelve months of the project, (03/01/95-02/29/96), we developed a second testbed, which we termed the ``Web prototype'' (WP), based upon available World Wide Web (WWW) technology;
3. during the past twelve months, (03/01/96-02/28/97), we have consolidated our testbed in many significant ways and, as a result, the system has been available for beta testing for some time now. Over 2000 individuals have become involved in this beta test program, providing valuable information for our current redesign efforts.

Several major activities are now taking place that will dominate many of our development activities during the coming twelve months (03/01/97-02/28/98). In particular,

1. after a six-month delay in the delivery of our main server and storage hardware, we are now ready to to set up public access to the Alexandria WP from WWW, and expect to go online in early 1997;
2. we have formulated a loading strategy and plan for an initial 100GB collection that will be loaded over the coming year together with the appropriate catalog information;
3. we have major activities underway that are leading a redesign and reimplementation of the testbed. We are basing our redesign on a model of a digital library as an extensible set of services that may be modeled in terms of distributed object concepts. We are also undertaking a major redesign and implementation of the user interface.

Much of the progress that we have made over the past year has been a result of a large number of fruitful interactions with partners from the private, public, and university sectors. For example, we have made and are continuing to make following efforts:

1. characterize the behavior of large databases in a new DEC hardware environment provided in part by DEC and the US Navy (NAVO, Stennis);
2. examine the applicability of ESRI's MapObjects as the basis for a map browser;
3. develop with Hughes a gazetteer that will be of use for EOS-DIS; we are partnering with Microsoft in the development of a 1TB online, Web-accessible database system;
4. work with NASA on the development of gazetteer content standards;
5. prototype distributed image access services based on the Common Object Request Broker Architecture (CORBA) in conjunction with NIMA;
6. apply and evaluate a variety of Oracle software packages, including relational and ``object oriented'' database software;
7. partner with SDSC in building a clone of ADL at SDSC and the embedding of digital library facilities in the next generation of supercomputing facilities;
8. work with USGS on a variety of data and metadata issues;
9. work closely with the US Navy (NRaD at San Diego in building high-performance Web servers.

Our research has progressed along many productive fronts during the past twelve months. As a result we now have a much better understanding of such issues as:

1. how to provide access to items in terms of ``fuzzy footprints'' and how to extract such information from a variety of sources;
2. semantic interoperability of metadata, particularly for images, and the development of visual thesauri;
3. user requirements in terms of access to geospatially-referenced materials;
4. the strengths and weaknesses of our interface design from the viewpoint of system users in our beta testing program;
5. many aspects concerning resource discovery and various issues relating to extensible data stores in digital libraries, such as multi-dimensional indexing, data placement, tertiary storage, content-based retrieval, and performance tuning;
6. image processing issues concerning access to geospatially referenced information, including wavelet transforms for both lossless and lossy compression, image segmentation algorithms utilizing texture and color information, and updating algorithms with applications to database search;
7. the construction of high-performance Web servers that are based on multicomputing concepts.

The Project has also developed a variety of other well-established activities that involve significant interactions with many groups outside of the Project. Apart from our close research and development interactions with the other DLI projects, we are about to hold our second Alexandria Design Review for many significant groups of users; our distinguished Advisory Board has just met for the third time; and we are beginning to develop significant interactions in the the area of K-12 education.

As noted above and in more detail in our Annual Project Plan, our plans for the coming year include:

1. the major redesign of both the testbed and the user interface in terms of distributed object technology and the loading of two major sets of collection items that will be publicly available.
2. Many extensions of our current research activities and the incorporation of the results of this research into our testbed development.
3. A continuation of our efforts in moving towards a truly operational digital library by the end of the project, in late 1998.