Edward A. Fox
Virginia Tech
Blacksburg, Virginia 24061-0106
http://www.ndltd.org
etd@ndltd.org
As worldwide activities in the area of digital libraries [FOX98] expand, and as discussions continue regarding international cooperation in this field [ACKS98], a key point of focus should be on education. Digital libraries containing resources that are helpful to learners can not only reduce costs and promote mutual understanding but also may speed up the transfer of knowledge and encourage international collaboration in the academy [FOX97b]. In the sections below we explore two such digital library initiatives, the Networked Digital Library of Theses and Dissertations (NDLTD) and the Computer Science Teaching Center (CSTC).
The Networked Digital Library of Theses and Dissertations (NDLTD) is open for all universities as well as other supporting institutions to join. As of June 1998 it included three national efforts, in USA, Australia, and Portugal. Member universities [NDLTDc] are drawn from Canada (including one group of three universities [TUG]), Germany, South Africa, and South Korea - with other members likely to join in scores of universities around the world. Thus, all universities in Russia are invited to join, as explained in the subsections below.
The first public meeting considering using electronic document processing methods, in particular SGML (Standard Generalized Markup Language), for theses and dissertations was held in Ann Arbor, Michigan, USA in November 1987. The meeting was hosted by UMI, and included representatives from the University of Michigan, SoftQuad, ArborText, and Virginia Tech (VT). Over the next year, VT funded and worked with SoftQuad to develop an SGML DTD (Document Type Definition) for theses and dissertations. This has been refined and used for a small number of dissertations, and is documented at a Virginia Tech WWW site [NDLTDf] which gives extensive instuctions so students can create Electronic Theses and Dissertations (ETDs).
An SGML-based form of each ETD is highly desirable, and should become feasible in the next few years as XML becomes popular and well-supported by producers of word processing software. That is one of the goals agreed upon in a 1994 meeting in Blacksburg, VA, USA attended by representatives of a number of universities interested in ETDs. The other goal, which became feasible by 1995, is submission of ETDs using Adobe's Portable Document Format (PDF), which affords a fully-rendered and directly readable version of student research results. Accordingly, with funding from the Southeastern Universities Research Association and from the U.S. Department of Education, the National Digital Library of Theses and Dissertations was launched [FOX96]. Since there has been extensive news coverage on this project [NDLTDb], some of which is misleading, and since the project has expanded into an international initiative, now called "Networked" instead of "National" [FOX97a], it is appropriate to explain the goals, problems faced, solutions and plans of NDLTD.
The goals of NDLTD include:
In addition to university-level goals, many students have their own goals regarding ETDs. Electronic documents can simply be the electronic version of a document designed for paper distribution, saving students the cost associated with photocopying and the hastles related to shipping if the student is not on campus but does have access to the Internet for file transmission. Of greater importance may be the goal of having works quickly and widely disseminated. Since ETDs often include a student's resume, and represent some of their finest work, rapid distribution may help students secure jobs or quickly become well known in their fields. This seems quite feasible, since electronic distribution may take place at the rate of hundreds of times more than by paper.
Even more exciting to some students is the possibility of an ETD being more expressive than traditional works [KIRS97]. Electronic documents may include color, other multimedia, hypertext links, data files, and other valuable resources usually only described. Students can upload files for network access that previously could only be shared on CD-ROM or diskette. Experience at VT indicates that roughly two-thirds of submissions have at least some multimedia. Thus, there are works with audio and video files, Java programs, a tutorial for the AuthorWare system, and VRML files for chemical structures.
Some faculty are particularly interested in improving the quality of theses and dissertations. The University of South Florida has an active writing center that hopes to boost the quality of the writing [USF]. Electronic tools to help diagnose and suggest fixes to problems, as well as systems that enhance collaboration with writing experts or scholars in the discipline, may be of value.
A key idea in NDLTD is to have students carry out the full activity, so they learn by doing. Besides preparing and submitting their ETDs to NDLTD, they also should learn about their intellectual property rights. Accordingly, VT ran a seminar in spring 1998 on scholarly publishing, with special emphasis on ethics, copyright, and other matters related to ETDs [VTRGS]. Audio files as well as transcripts are available from the series, for individual or group use at other institutions. It is hoped that other institutions will undertake similar public discussions, since moving to ETDs has opened up a broad range of little-considered concerns.
Some of the most challenging problems associated with digital libraries relate to social or cultural concerns [BORG96]. Since our aim with NDLTD is to have a profound effect on scholarly publishing, university sharing of information, dissemination of knowledge, and graduate education, it is not surprising that people have voiced concerns or identified problem areas.
One set of issues deals with variations between cultures and languages [BORG96]. Since theses and dissertations are written by graduate students worldwide, usually in their native language, often describing or set in the context of local culture, and can more effectively carry their message when multimedia content is included, ETDs will vary regarding language, cultural portrayal, and use of multimedia forms. Handling such works presents unique challenges regarding cataloging, indexing, search, storing, distributing and preserving ETDs. We welcome these challenges, since digital libraries must meet them. Solutions seem feasible in all cases, as is explained below.
The digital library community has effected solutions to most of the problems raised regarding ETDs. There are systems for multilingual searching and mechanisms for handling multimedia information. Universities involved in NDLTD adapt the WWW, automation, and training materials developed by VT to suit their local culture, policies, and procedures. Any of a large number of search tools can be deployed by a participating institution so it can become part of the simple federated search system developed at VT and accessible at the central collection site [NDLTDa].
More serious concerns arise in connection with publishers. We argue that ETDs are a separate and unique genre, that should operate independently of other types of publications. Thus, an ETD is a much longer and more detailed than a paper that might appear in a conference proceedings or journal. Similarly, a humanities dissertation usually goes through substantial refinement and enhancement before appearing through a university press as a monograph.
The first solution we have adopted regarding confusion about affording widespread access is to work with publishers. We have spoken at a number of conferences and participated in many meetings to explore this topic. Through our training materials and workshops we help educate students about the broad area of intellectual property rights and the specific issues relating to copyright and right of first publication.
Four publishers have provided policy letters, which we have scanned and made available at our WWW site, so students understand their conditions on making works available through NDLTD while at the same time publishing derivative articles in the journal literature. We hope that other publishers will provide similar statements which we also can make available so large numbers of students will not have to contact individual editors to discuss this issue.
Second, we developed an Approval Form [NDLTDg] that is signed by students and their faculty committee. It allows them to:
The Approval Form protects students from being penalized by publishers. However, it is very easy to just restrict access by checking a box, which then substantially delays or reduces the number of people who will read an ETD. Further work with publishers on policies, and social pressure from the increasing number of universities in NDLTD to allow ETDs to be more freely shared, is needed.
TO BE WRITTEN
[GLAD97] document access control [LYNC97] [PAYE97] [SCP]etd home page [UMI]
[NDLTDe] related projects For more information about NDLTD, the reader is referred to an online bibliography and to copies of key papers [NDLTDd].
The Computer Science Teaching Center (CSTC)
OLD STUFF ================================================================ there are tangible results of this work: an emerging infrastructure, a set of enabling tools, and a collection of content and resources.
When a university joins NDLTD, it begins to develop the infrastructure to support its students as they learn about electronic publishing, digital libraries, intellectual property rights, and other key concerns necessary for them to function effectively in the Information Age. This often includes a small WWW site, typically adapted from the pages in use at Virginia Tech [NDLTDf]. If students use the online WWW-based submission software provided by Virginia Tech to upload the files that make up their ETDs, the university may support that as well as storage of the files while their work is being checked for conformity with graduate policies and is being cataloged for library handling. If the university elects to itself make ETDs available, some type of digital library support is needed too.
Probably the most elaborate infrastructure involved in the ETD initiative is provided by institutions encouraging the use of multimedia content in theses or dissertations. Devices for capturing, converting and manipulating images, video, audio, virtual reality environments, etc. can be provided by libraries, computing organizations, or in a variety of places around each campus. At Virginia Tech this is handled through the New Media Center in the Library as well as in a number of computer laboratories.
In addition to hardware and software to support NDLTD, Virginia Tech also has a rich network infrastructure, including a vBNS (high speed Internet research and education backbone) connection through "Network Virginia," the statewide ATM network with over 220 nodes that it runs and which includes educational institutions all over the Commonwealth. This connection ultimately links to SINGAREN, to reach Singapore and beyond with high speed networked communications. Since Virginia Tech has offered to provide archival support for NDLTD members in the short term, through an IBM-donated server with four terabytes of storage, it is useful to have such comprehensive infrastructure, though that is certainly not required for NDLTD members.
Indeed, since UMI has a business activity that provides preservation and access services for theses and dissertations, involving paper, microforms, and now electronic services, institutions can elect to just support student submission and local approvals, paying UMI to handle other matters [UMI]. The NDLTD Steering Committee is pleased that both UMI and OCLC have indicated interest in long-term archival support, and encourages an open competitive environment in which a variety of efforts evolve to maintain and expand access to ETDs.
Tools have been developed, and more are needed, to effectively support graduate education. Students need word processing software to create their documents. Adobe's Acrobat products support the handling of Portable Document Format (PDF) files, the format usually chosen for ETD submission. Use of SGML is encouraged, aided by Virginia Tech's document type definition (DTD) for ETD-ML, the markup scheme which has been repeatedly refined to be easy to use and yet powerful enough to capture the important metadata and structure of ETDs. Efforts are underway at the University of Virginia and the University of Michigan to apply the guidelines of the Text Encoding Initiative and their DTDs for students willing to include much more extensive markup. Virginia Tech will have guidelines by 1999, when XML solutions will become feasible for large numbers of students, so authors working with Microsoft Word and other packages can export their ETDs into that emerging archival form.
For universities wishing to afford access to their ETD collections, local search engines are of value. Virginia Tech used the Free WAIS software initially, which is adequate for the task, but then shifted to the more powerful OpenText system which was licensed for general searching efforts around campus. OCLC has donated a copy of their SiteSearch software, which can provide another alternative. SiteSearch supports Z39.50, which enables access through a variety of clients [LYNC97]. It also can be adapted to enable "federated search" of all ETD repositories in parallel, with client or gateway merging of results from remote sites[PAYE97].
The University of Virginia has taken the initiative on adapting the Dienst system (developed at Cornell, and used in another federated search environment, the Networked Computer Science Technical Report Library, <http://www.ncstrl.org>) to use for ETDs. Work is underway at Virginia Tech to use IBM Digital Library as yet another alternative to afford access; in this case there is strong support for protecting intellectual property [GLAD97].
As part of the research investigations related to NDLTD, there have been a variety of efforts exploring new types of tools and evaluating existing ones. For example, in Virginia Tech course CS6604, Digital Libraries, taught in the fall of 1997, a large team developed a virtual reality library to house the ETDs, with different floors for the various disciplines (e.g., Engineering or Business), different rooms for each department, and both bookshelves for the ETDs as well as murals for images extracted from them. Another group of students in the class, who previously completed a course on human-computer interaction, undertook a careful controlled study of four digital libraries: ACM, IEEE-CS, NCSTRL, and NDLTD. Suggestions emerged regarding making the interface clearer for novice users. Pre- and post-surveys showed that users changed their expectations for the digital library systems after several hours of use, indicating that it is hard to solicit requirements and design objectives unless users are engaged in an ongoing formative evaluation effort.
Since there are many places in which tools and other resources play a role, Virginia Tech's WWW materials related to ETDs have been reorganized into three parts, to discriminate clearly among the following:
At Virginia Tech, pilot work toward NDLTD has been underway for more than 10 years. A number of theses were converted by hand from word processing formats into SGML, and later into PDF when it became available. In 1996, students submissions were encouraged in SGML, PDF and LaTeX (to be converted in the Graduate School to PDF). Virginia Tech began requiring electronic submission at the start of 1997, so the collection size went from around 80 at the end of 1996 to over 500 by the end of 1997, to over 1100 by June 1998. About two-thirds of these have multimedia content, mostly color images, but some have audio, video, or interactive multimedia components as well.
At other universities similar growth is starting, but it is delayed slightly because of when they joined NDLTD and gated by how energetically they have promoted the effort. Small collections exist at a number of institutions, and by June 1998 there were five searchable university collections linked from the project collection home page.
While there were about 37K accesses to the Virginia Tech collection in 1996, there were 248K in 1997 and 218K in January-May 1998. That amounts to a growth in average daily rate from 102 to 685 to 1432. The number of PDF files downloaded (approximately equal to the number of full ETDs downloaded, since most are made up of a single PDF file) rose from about 5K in 1996 to 73K in 1997 to 118K in the first part of 1998. There were 9K different sites served in 1996, 23K in 1997, and already in 1998, 17K. Access statistics are made available through Virginia Tech's Library [SCP], and demonstrate substantial interest worldwide. Accordingly there are a broad range of activities underway to expand NDLTD and its services.
Though enormous progress has been made in NDLTD in about a year and a half, further expansion by an order of magnitude is needed for comprehensive handling of the hundreds of thousands of graduate theses and dissertations produced each year.
First, to extend the collection, there must be more complete coverage of the universities involved. While NDLTD members that are very large, like University of Wisconsin at Madison and University of Texas at Austin, have many graduate students who will prepare theses, and it is important to involve such institutions, our ultimate goal is to serve the needs of all interested institutions and students. Accordingly, hundreds of universities have been invited to join NDLTD, through visits, presentations, publications [NDLTDd], mailings, and other mechanisms. In addition, since initially membership was restricted to universities that agreed to eventually require their students to submit ETDs (which is the most effective way to guarantee large numbers of submissions quickly, and which leads to considerable savings in campus effort as opposed to what is required to handle both paper and electronic versions), new classes of membership have been created to further stimulate involvement.
Type 1 members require ETDs - there are 2 universities in this category. Type 2 members agree to require ETDs at some point in the future. Other types of membership encourage:
Expanding NDLTD has been aided by the extensive publicity afforded by TV, radio, and the press [NDLTDb], which realize that requiring ETDs is a very significant step in the arena of education and applications of technology, as well as evidence of a new type of involvement of universities in publishing. In part due to lack of understanding about access issues, the press has perceived the move toward ETDs to be relatively controversial.
If a student wishes to have an ETD accessible to the world, and their graduate committee agrees, they indicate that on their ETD approval form [NDLTDg], which serves as the agreement with the university regarding submission and access. Most of the Virginia Tech collection is in that category. Our most popular works have been downloaded thousands of times, indicating the enormous demand for detailed research studies.
On the other hand, students seeking patent protection can use digital library technology to lock up their works, typically for a year. Alternatively, and this is done primarily to resolve any concerns of publishers, they may restrict access to the campus. In the case of ACM and IEEE-CS, associations serving the computing field, as well as Elsevier, the largest scientific and technical publisher, NDLTD has solicited and received letters saying that such restriction is not necessary. Thus, these publishers do not consider that making an ETD available constitutes prior publication of what might be sent to them for a journal or conference paper. We are working actively to have other publishers provide similar statements.
Another way to increase access is to make such access easier for students, researchers and other scholars. Thus, at the Virginia Tech WWW site, in addition to searching with OpenText one can browse. Later, other search systems will provide additional support. At the level of NDLTD, though, a very important service provided is multilingual federated search.
James Powell has developed a simple federated search system for Virginia Tech. Used originally only for campus servers, it now is also in place for NDLTD. It lets users provide a query that can be matched against site descriptions, to aid in selected suitable sites to search. Once one or more sites have been selected, or in the default case when all are selected, a full query is sent in parallel to those sites. Since the site descriptions specify the language of the local collection, the federated system translates words in the query to other languages as needed. This occurs in NDLTD for the case of a site in Austria, which uses German.
One of the sites involved in the federated search is at the University of Virginia, running Dienst. Thus it is possible for NDLTD to include sites running Dienst, or for a regional service (e.g., one coordinated by the Portugese National Library for universities around the nation) to be a part of the overall NDLTD umbrella. Since Virginia Tech has participated on the NCSTRL Working Group since its inception, further discussions about the interplay between NDLTD and NCSTRL will be explored in upcoming months.
Of key importance to NDLTD members is handling preservation. Corporate support for theses and dissertations by UMI emerged in part from the perception that these works have long-term archival value and should be preserved for fifty, seventy, one hundred, or more years. Many public universities have a legal obligation to preserve these documents for at least fifty years. Since Virginia Tech has such an obligation, and only accepts ETDs, not paper documents, this is a serious matter, for which the Library has assumed responsibility.
One approach is to delegate such responsibilty to a trusted third party, such as UMI or OCLC (which offers preservation services in connection with its Electronic Collections Online product). Another approach is for groups of universities, e.g., the CIC, to undertake such services in a centralized fashion to benefit from economy of scale.
In addition to these logistical issues, there are the technical problems and corresponding solutions. Virginia Tech has opted to keep its ETD collection online, rather than focusing on a storage solution involving CD-ROM (selected by UMI) or tape. But migrating to new media as technologies shift is only part of the problem. More serious is the problem of changes in electronic format. The choice of NDLTD universities to work with PDF and SGML is based on concensus in the community that an open international standard like SGML should be effective for preservation, and assurances from Adobe that PDF will continue to be supported and have a published specification, and that all new versions will be understood by freely available reader software provided for all common platforms, which will be able to handle all prior versions as well.
Regarding use of other formats by students, NDLTD provides a list of recommended standards that are likely to be easy to preserve. It also is explained that use of other formats can be done at the discretion of students and their committees, when such use is necessary for effective communication, but may lead to parts of an ETD becoming unusable in a small number of years.
Most problematic in this arena, however, is the issue of economics. While universities do collect some fees from students for each ETD, often on the order of $20, to cover archiving (formerly, a "binding fee"), it is not clear how much will be required to fund access and preservation over the lifetime of an ETD. At present, the quantity of data (usually 1-2 megabytes per ETD, though including multimedia can increase the size by one or two orders of magnitude) is small, and since few universities will have more than about a thousand per year, storage costs seem minimal. While accesses of thousands per day are impressive relative to the 2-3 circulations per year of paper theses, even PC-scale servers can handle several orders of magnitude more. Thus, it may be that the dominant cost for participation in NDLTD relates to preservation. Since NDLTD is aiming to lower the processing costs for ETDs through end-to-end automation, which appears to have been quite succesful according to reports at member institutions, and since many studies suggest that electronic preservation costs over the long term will be less than for traditional books (also, less than the roughly $50 charged by UMI), it seems plausible that an economically viable solution will evolve. Indeed, since much of university knowledge is in electronic form in general, and preservation of that must be effected, it is probably fortuitous that these serious concerns receive attention through NDLTD, where collaboration and technical investigations are the watchword, and where the quantity and complexity of materials is relatively simple (relative to other sources like courseware, data sets, lab notebooks, and general records).
Another important activity in NDLTD is training of students so they are empowered to create electronic documents and to use digital libraries. Toward this end, the NDLTD WWW site contains hundreds of megabytes of tutorials, checklists, files of frequently asked questions (with answers), and presentations that can be adapted for training workshops. At Virginia Tech, there are open workshops run at least six times per year for students to learn about creating ETDs. Based on analysis of surveys from those meetings, it became clear that one set of workshops was needed for the average student, and another type of presentation was required for those with complex requirements. In addition, there must be a small pool of people somewhere on campus to help students who have trouble when finalizing their ETD. These helpers can assist not only those needing advice about scanning, but also those having trouble with the "basics".
Though there are extensive computer literacy efforts at Virginia Tech, nevertheless there are many students preparing to receive graduate degrees who have minimal knowledge even of fundamental word processing concepts. Often students do not understand about fonts, figures and tables. Besides having trouble using common tools, they have difficulty with the transfer of diagrams and other components between tools, failing to distinguish between graphic objects and bitmap representations until they find that paper versions of diagrams have poor quality whereas their on-screen originals seem perfect. Many graduate students would never face or resolve these problems unless they arise in the course of their own ETD preparation. Since students often learn best by doing, this competency-oriented program should ensure that the next generation of scholars is prepared more completely for the Information Age.
Evaluating the progress of NDLTD involves a number of activities. At the highest level, there is the monitoring of efforts at a far-flung group of universities. Email and WWW surveys are helpful, but greater bandwidth in reporting resulted from a user-group style meeting held in Memphis in June 1998. At that point it was decided that the regular meetings held at spring and fall CNI events would be shifted from sessions of the NDLTD Technical Advisory Committee to a more inclusive member discussion.
At the collection level, there is online monitoring of the use of the collection. An optional WWW form is filled in by users of the OpenText system that supports searching of the Virginia Tech collection. While the pilot form needs refinement, and supplemental collection schemes will be required, there are some interesting preliminary findings. First, it appears that most users preferred searching as opposed to browsing, even in this small collection of 1100 works. Second, of the abstract files viewed, roughly 10% were reported as downloaded while 20% were printed. Third, since the collection is small, slightly more accesses to other collections about theses and dissertations were reported than to the Virginia Tech collection, with roughly a quarter of the total to UMI's services, and roughly 10% of the total leading to interlibrary loan.
Fourth, regarding publishing plans, about 15% indicated having already published in a journal, two-thirds planned to, and about a quarter planned to prepare a book. Most with publishing plans expect to complete that within a year; roughly 30% expect to take longer. Worry about not being able to publish if an ETD is made world accessible continues to concern almost half of the respondents, causing them to restrict access. Only a very small number considered that affording different levels of access to different parts of an ETD made sense.
Other evaluation efforts continue regarding surveys about NDLTD given to students as they complete their degree requirements, and another instrument used to get comments on the ETD submission process. Reporting of these findings will proceed, especially in periodic reports to U.S. Department of Education and the Steering Committee, which will be available on project WWW pages. All of these results will be influenced by future growth and development of NDLTD.
Based on experience with NDLTD, it appears likely that further development will be of considerable benefit to students, universities, researchers and other scholars by making an important genre widely available. Doing so depends upon three key future types of growth.
First, many governments and other institutions (e.g., in U.S., Singapore) have argued that there are important requirements to enhance education, and that technological solutions, including digital libraries, may be promising mechanisms. It is clear from the experiences with NDLTD that:
In addition to these practical and skill-oriented observations, there are other less tangible benefits relating to objectives of students understanding more about electronic publishing. Since many students elect to restrict access to their works out of concern about publishers, there is at least some thought about intellectual property, policies regarding prior publication, and the effect that access has on publication options. To increase understanding about these many issues, Virginia Tech ran a seminar in spring 1998 with invited speakers discussing copyright, intellectual property rights, ethics, and other topics relating to ETDs. These are made available through digital audio and online transcripts for others to explore [VTRGS].
Further enhancement to graduate education may results from additional work related to NDLTD. An undergraduate honors thesis is underway to define requirements and implement a prototype environment to support graduate student research, including access to NDLTD. Other efforts will proceed, given current support, and may be extended further if more proposals are funded.
Another future benefit of NDLTD is its influence in helping universities work together. All research universities have theses and dissertations, and smaller institutions have bachelor theses and other major papers. These parts of the "grey literature" are largely ignored and disappear from view, except in the case of about fifty thousand dissertations in North America. With hundreds of thousands of student involved in the creation of such documents, often spending months or years in preparing them, there is strong motivation for making them more accessible. With significant faculty time and other university resources going into the development of these works, there is an incentive at the university level to proceed with NDLTD, especially since it ensures some basic level of electronic publishing skills among graduates, and reduces overall costs.
With these motivations in place, and with a number of universities interested in helping others ensure that ETDs increase in quality as well as quantity (e.g., through writing enhancement and inclusion of multimedia components), it appears that there are incentives for universities to collaborate.
Since there are few other venues in which universities worldwide can cooperate and all benefit, and since working on NDLTD will ensure that participating universities develop at least a basic infrastructure in the digital library arena, it appears that NDLTD may be a good vehicle for both university cooperation at the international level and a means for the digital library field to advance.
Finally, it should be noted that since the digital library field is new, and since our systems and services are still largely prototypes, there is considerable room for evolution. This actually fits in well with the development of NDLTD, which is naturally gated by the rate of joining, the training efforts on local campuses, and the implementation of services by different institutions. Since universities can start small, and since the collection is still in its early growth phase, there is time for additional funding and research investigations to be launched. This is valuable, since requirements are just becoming obvious, and prototypes as well as evaluation efforts are in their early stages.
It is clear that we have many challenges, regarding training students about digital libraries, working with publishers to allow an increase in access, developing better archiving mechanisms (e.g., using XML), enhancing search services (with better multilingual, multimedia, and federated search capabilities). Many of these improvements will arise more rapidly if there is greater involvement from the East in this clearly global enterprise.
[AKSC98] Akscyn, R. and Witten, I. (1998). "Summit on International Cooperation on Digital Libraries." Report of ACM Digital Libraries Workshop; June 27-28, 1998; Pittsburgh, PA; <http://www.ks.com/dl98>
[BORG96] Borgman, C.L.; Bates, M.J.; Cloonan, M.V.; Efthimiadis, E.N.; Gilliland-Swetland, A.; Kafai, Y.; Leazer, G.L.; Maddox, A. (1996). "Social Aspects Of Digital Libraries." Final Report to the National Science Foundation; Computer, Information Science, and Engineering Directorate; Division of Information, Robotics, and Intelligent Systems; Information Technology and Organizations Program. Award number 95-28808. <http://dlis.gseis.ucla.edu/DL/>
[BORG97] Christine L. Borgman (1997). "Multi-Media, Multi-Cultural, and Multi-Lingual Digital Libraries: Or How Do We Exchange Data In 400 Languages?" D-Lib Magazine, June 1997.<http://www.dlib.org/dlib/june97/06borgman.html>
[FOX96] Edward A. Fox, John L. Eaton, Gail McMillan, Neill A. Kipp, Laura Weiss, Emilio Arce, and Scott Guyer (1996). "National Digital Library of Theses and Dissertations: A Scalable and Sustainable Approach to Unlock University Resources." D-Lib Magazine, September 1996.<http://www.dlib.org/dlib/september96/theses/09fox.html>
[FOX97a] Edward A. Fox, John L. Eaton, Gail McMillan, Neill A. Kipp, Paul Mather, Tim McGonigle, William Schweiker, and Brian DeVane (1997). "Networked Digital Library of Theses and Dissertations: An International Effort Unlocking University Resources." D-Lib Magazine, September 1997.<http://www.dlib.org/dlib/september97/theses/09fox.html>
[FOX97b] Edward A. Fox, Robert Hall, Neill A. Kipp, John L. Eaton, Gail McMillan, and Paul Mather (1997). "NDLTD: Encouraging International Collaboration in the Academy." DESIDOC Bulletin of Information Technology, September 1997. <http://www.ndltd.org/pubs/dbit.pdf>
[FOX98] Edward A. Fox and Gary Marchionini (1998). "Toward a Worldwide Digital Library. Guest Editors' Introduction to special section (pp. 28-98) on Digital Libraries: Global Scope, Unlimited Access." Commun. of the ACM, Apr. 1998, 41(4): 28-32. <http://purl.lib.vt.edu/dlib/pubs/CACM199804>
[GLAD97] Henry M. Gladney (1997). "Safeguarding Digital Library Contents and Users: Document Access Control." D-Lib Magazine, June 1997.<http://www.dlib.org/dlib/june97/ibm/06gladney.html>
[GUED98] Jean-Claude Guédon (1998). "The Virtual Library: An Oxymoron?" NLM and MLA 1998 Leiter Lecture, National Library of Medicine, Bethesda, MD, May 1998.
[KIRS97] Matthew G. Kirschenbaum (1997). "Electronic theses and dissertations in the humanities: A directory of on-line references and resources."<http://etext.lib.virginia.edu/ETD/ETD.html>
[LYNC97] Clifford A. Lynch (1997). "The Z39.50 Information Retrieval Standard: Part I: A Strategic View of Its Past, Present and Future." D-Lib Magazine, April 1997.<http://www.dlib.org/dlib/april97/04lynch.html>
[NDLTDa] NDLTD Team (1998). "ETD Digital Library."<http://www.theses.org/>
[NDLTDb] NDLTD Team (1998). "NDLTD in the News."<http://www.ndltd.org/news/>
[NDLTDc] NDLTD Team (1998). "NDLTD Official Members."<http://www.ndltd.org/members/>
[NDLTDd] NDLTD Team (1998). "NDLTD Papers and Publications."<http://www.ndltd.org/pubs/>
[NDLTDe] NDLTD Team (1998). "NDLTD Related Projects."<http://www.ndltd.org/related/projects.htm>
[NDLTDf] NDLTD Team (1998). "Virginia Tech Electronic Thesis and Dissertation home page."<http://etd.vt.edu/>
[NDLTDg] NDLTD Team (1997). "Virginia Tech Graduate School Electronic Submission Approval Form."<http://etd.vt.edu/submit/approval.htm>
[SCP] Scholarly Communications Project (1998). "Scholarly Communications Project: Virginia Tech Electronic Thesis and Dissertation home page."<http://scholar.lib.vt.edu/theses/>
[PAYE97] Sandra D. Payette and Oya Y. Rieger (1997). "Z39.50: The User's Perspective." D-Lib Magazine, April 1997.<http://www.dlib.org/dlib/april97/cornell/04payette.html>
[TUG] TriUniversity Group (TUG) (1998). "TUG Electronic Thesis Project."<http://www.lib.uwaterloo.ca/TUG/ETD/>
[UMI] UMI. (1998). "ProQuest Digital Dissertations."<http://wwwlib.umi.com/solutions/2.0.html>
[USF] University of South Florida ETD Project (1998). "Electronic Publication of Theses and Dissertations, University of South Florida."<http://www.usf.edu/~writing/etds.html>
[VTRGS] Virginia Tech Research and Graduate Studies (1998). "New Issues in Academe: Scholarship in the Electronic World."<http://www.rgs.vt.edu/resmag/seminars.html>
Many thanks go to faculty, students and staff at Virginia Tech and at other institutions who have worked on ETDs, especially John Eaton, Gail McMillan, Neill Kipp, Paul Mather, Robert Hall, Bill Schweiker, and Todd Miller. The U.S. Department of Education's Fund for the Improvement of Post Secondary Education supports NDLTD. Additional in-kind support has been provided by many parties including: Adobe, Arbortext, Council of Graduate Schools, CNI, IBM, Microsoft, OCLC, SOLINET, and SURA.
The Computer Science Teaching Center is funded by NSF and the ACM Education Committee, and is run with co-PIs Deborah Knox of The College of New Jersey and Scott Grissom of the University of Illinois, Springfield. NSF also funds Curriculum Resources in Interactive Multimedia, with co-PI Rachelle Heller of The George Washington University.