Report library offers many benefits

By Alan L. Selman
CRA Staff

(The Computing Research Association strongly endorses the concept of unified electronic access to computer science technical report literature and encourages member institutions to catalog their reports using the Networked Computer Science Technical Report Library [NCSTRL] system. At its July meeting, the CRA Board passed a motion supporting NCSTRL.)

The Networked Computer Science Technical Report Library (NCSTRL) is a digital library for technical reports from computer science departments and research laboratories. This library will benefit the departments that contribute to it, researchers who consult it, students who access the library for their course work and authors whose work makes up the corpus.

Currently, 29 Computer Science Departments and research laboratories contribute their department technical reports to NCSTRL. To provide the level of service that is intended, all Ph.D.-granting departments and research laboratories need to join. This article represents the start of a campaign to urge your department to contribute your technical reports to this digital library. You will find that NCSTRL is stable and provides convenient, quality access. Minimal effort is required of departments, and the cost is far less than the cost of managing a paper collection.

It is becoming more difficult for departments to manage their technical report collections. An increasing number of high-quality reports are being produced at more research organizations, yet department operating budgets are decreasing. By contributing to NCSTRL, departments will eliminate the cost of printing, storing and distributing paper technical reports. Charging for reports would merely escalate costs throughout the research community.

Similarly, researchers need greater access to research material. Young researchers, for example, who hold faculty positions at colleges and universities that do not have strong research traditions will depend on NCSTRL for access to the latest research results. Authors will receive the widest exposure of their works, which will be searchable worldwide. Students at the undergraduate and graduate levels will have easy access to research results not yet available in journals and conference proceedings. New and prospective graduate students can find out about the work of faculty electronically, with minimal guidance. Thesis students can use NCSTRL to familiarize themselves with current research in their chosen area.

NCSTRL has a World Wide Web interface. The interface supports searching by author, title or abstract, and results may be browsed by table of contents or thumbnail images, read on the screen or printed.

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

1) 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.

2) 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 presents a hit list to the user.

3) 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 online or for downloading for printing.

The technology underlying NCSTRL is a network of interoperating digital library servers and FTP repositories. A department may choose to contribute in one of two ways: either as an NCSTRL-Standard site or as an NCSTRL-Lite site. A Standard site runs a server and maintains an FTP repository; a Lite site maintains only an FTP directory. The level of functionality for Standard sites is higher than for Lite sites. To make this choice, an individual department needs to balance the level of resources it can dedicate with the level of functionality it wishes to provide.

There are several differences in functionality:

• Standard sites can provide customized user interfaces to their collection. They can create special browsing interfaces to help users scan all documents in the local collection. Lite sites do not run any local user interface and rely on one of the other user interface gateways.
• 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. Lite sites submit bibliographic records to the central indexing site and do not control any of the actual indexing technology.
• Standard sites have the ability to store and provide access to documents in multiple formats. Lite sites can only provide access to a single document format, generally PostScript.
• Standard sites can provide a number of additional user interface features, such as page zooming, which is an interface for downloading and printing selected pages, and an interface for displaying and accessing the logical structure of documents--chapters, sections and so on.

The cost of sites are the start-up expenses of installing the software and preparing bibliographic records for the initial collection and the continuing costs of installing new releases and maintaining the collection. Lite takes about an hour to install, and Standard takes about two days. E-mail and telephone assistance will be available. Preparing bibliographic records is a simple clerical task. Installing new releases takes about two days a year. The largest cost is establishing the procedures to ensure that new reports continue to be entered. The actual labor is small, only a few minutes per document, but it requires a commitment from every author in the department. The cost of storage is small; a high-capacity disk can easily store all of a site's reports (in PostScript). Many departments already make all their reports available by FTP. These departments will find they can increase accessibility with a minimal increase in cost or effort. Most sites will place only their recent reports online, while others may choose to place their entire collection online, which will entail scanning their collection of older paper reports. A department might initially choose Lite, then move up to Standard after gaining experience.

The NCSTRL project and architecture is the result of a collaboration of two earlier successful technical report services. The first of these is Dienst, a result of the Advanced Research Projects Agency-funded CS-TR project. The second is the National Science Foundation-funded Wide Area Technical Report Service, which originated as a recommendation of a workshop on managing department technical reports at the 1992 CRA Conference at Snowbird. Stable organizational structure will be provided by CRA representation and by the NCSTRL Working Group. The latter is an activity of the DLIB Forum, which is supported by ARPA. The DLIB Forum is sponsored by the Information Infrastructure Technology and Applications Task Group of the High-Performance Computing and Communications program.

We are eager to add your department to the library. We stress that NCSTRL is self-explanatory and easy to use. Anyone is welcome to search and read the documents, provided they respect the terms and conditions of the contributors. To try NCSTRL, learn more about NCSTRL or get detailed installation information, see the NCSTRL home page at http://www.ncstrl.org.

Alan L. Selman is professor and chair of the Department of Computer Science at the State University of New York at Buffalo. His principal research interest is complexity theory. Selman was one of the collaborators who helped to formulate WATERS, a precursor of NCSTRL.