NCSTRL Technical Overview

The Networked Computer Science Technical Report Library (NCSTRL) is a distributed system providing a single point of access to research results from international computer science departments and laboratories. Researchers throughout the world aree able to use familiar Internet tools to search for, browse, read, and download technical reports from participating institutions.

The NCSTRL project and architecture is based on Dienst, a result of the ARPA-funded CS-TR project, and the NSF-funded Wide Area Technical Report Service (WATERS). NCSTRL will continue to evolve in parallel with national digital library research efforts. Future versions of NCSTRL will be based on a distributed digital object infrastructure as articulated by Robert Kahn of CNRI and Robert Wilensky of UC Berkeley. Compatibility with earlier versions of NCSTRL will be maintained. NCSTRL will provide a working model of this infrastructure, allowing members of the NCSTRL community to test its effectiveness and participate in its continuing development.

NCSTRL users search for and retrieve documents from a unified collection

From the user's point of the view NCSTRL appears as a single collection even though, physically, the repositories containing the collection and the engines for searching over it are distributed over the Internet. A user wanting to find a document in the distributed collection follows a simple procedure:

• Use a standard WWW browser to open a connection to one of the user interface gateways to the NCSTRL collection. The gateway returns a simple form for searching the collection.
• Complete the search form with criteria for a search of the collection, for example ``author = hopcroft'', and submit the search. The gateway processes the search and present a hit list to the user.
• Select one of the documents from the hit list. The gateway retrieves the document and makes it available to the user for reading or browsing on-line, or downloading for printing.

A technical infrastructure based on interoperating distributed servers

The technology underlying NCSTRL is a network of interoperating digital library servers and FTP repositories. The digital library servers interoperate using an open protocol and provide three services; repository services that store and provide access to documents, index services that allow searches over bibliographic records, and user interface services that provide the human front-end for the other services. Search requests from users generate parallel protocol requests to the distributed index servers. A distinguished central index server acts as the search engine for the documents that are stored in FTP repositories. Document access requests generate a protocol request to the repository in the network that contains the document.

The distinction between the central or distributed indexes and between FTP repositories or NCSTRL repositories is invisible to the user. All documents from all institutions are searchable through any of the user interface gateways. The results of searches are selectable links to the repository copies of documents that match the search criteria. The interface for documents resident in a NCSTRL server repository or an FTP repository site is exactly the same.

Interoperation of digital library services through an open protocol is an important aspect of the system. It allows access to the collection and its indexing data not only from the user interface services, but from other value-added services. These may be other more advanced indexing systems, information retrieval tools, natural language analyzers, and the like. We foresee that the open architecture of the NCSTRL will make it a significant corpus for a variety of research projects.

How does an institution participate in NCSTRL? Lite or Standard?

An institution may choose to be an NCSTRL-standard site, in which case it runs a server, or an NCSTRL-lite site, in which case it only maintains an FTP repository. To make this choice an individual institution should examine at least the following two questions.

What are the hardware and personnel resources that the institution wishes to dedicate to NCSTRL access? Many institutions already store technical reports in FTP repositories and simply want to link these repositories to a larger collection. Other institutions have experimented with higher levels of access, such as local WAIS indexes or WWW pages, and want to combine high functionality with interoperability.

What level of functionality does the institution want to provide for researchers that access their collection?

• NCSTRL-standard sites can provide customized user interfaces to their collection. For example, they can create special browsing interfaces to help users scan all documents in the local collection. NCSTRL-lite sites do not run any local user interface and rely on one of the other user interface gateways.
• NCSTRL-standard sites store bibliographic data locally and maintain their own search engines. This gives these sites immediate control over the contents of bibliographic records and the ability to install more powerful search engines and search interfaces. NCSTRL-lite sites submit bibliographic records to the central indexing site and do not control any of the actual indexing technology.
• NCSTRL-standard sites have the ability to store and provide access to documents in multiple formats. GIF images allow users to read documents on-line. Thumbnail page images allow users to browse a document to find graphs or tables. High-resolution TIFF images are useful for users who want the best quality archival version of a document. NCSTRL-lite sites can only provide access to a single document format, generally PostScript.
• NCSTRL-standard sites can provide a number of additional user interface features such as page ``zooming'', an interface for downloading and/or printing selected pages, and an interface for displaying and access the logical structure of documents (chapters, sections, etc.).

Acknowledgements

This work was supported in part by the Advanced Research Projects Agency under Grant No. MDA972-92-J-1029 with the Corporation for National Research Initiatives (CNRI) and by the National Science Foundation under Grant No. NSF-CDA-9308259. Its content does not necessarily reflect the position or the policy of the Government or CNRI, and no official endorsement should be inferred.

Up to Main Information Menu

NCSTRL Documentation
Any comments or questions?
Contact us at help@ncstrl.org