Introduction

In March of 2006, the State Library of North Carolina hosted Digital Preservation in State Government: Best Practices Exchange 2006. [1] This rich two-day forum built upon the efforts of the The National Digital Information Infrastructure and Preservation Program (NDIIPP) by facilitating communication between librarians, archivists, and other information professionals. Over one hundred participants from thirty states, federal agencies, and private industry attended the Exchange, allowing for open communication and idea sharing between traditionally autonomous institutions.

The Best Practices Exchange was made up of two facilitated large-group sessions (an opening session and a closing wrap-up) and thirty small-group exchange sessions that focused on one of nine topics. Multiple sessions were conducted simultaneously throughout the two-day event, allowing for more intimate “round table” meetings where all attendees could feel comfortable participating. Designated contributors gave informal presentations where they shared their experiences with managing digital information. These presentations inspired questions and discussions among the other attendees, allowing for a far more participatory experience than a typical conference.

Each exchange session focused on one of nine aspects of digital preservation: identification, selection, and appraisal of digital assets; repository systems; collection of digital assets; authentication; resources and workflows; access; metadata; preservation; and organization. Staff from the State Library of North Carolina or the North Carolina State Archives facilitated and recorded each exchange session. Additionally, visual journalist Eileen Clegg graphically represented the major themes and ideas generated throughout the Best Practices Exchange on large wall murals.

Figure 1. These images from the murals of visual journalist, Eileen Clegg illustrate some overarching ideas discussed at the Best Practices Exchange.

Examining the recordings and wall murals from the Best Practices Exchange reveals a common message that runs throughout all the sessions. Participants stressed again and again that a successful digital preservation program requires a strong foundation. During discussions, one participant compared the digital environment to a desert with constantly shifting sands. A strong foundation would allow the digital preservation program to resist the shifting sands and remain stable in a highly dynamic environment.

Participants identified four essential elements for building a strong foundation for a digital preservation program: support and buy-in from stakeholders; “good enough” practices implemented now; collaborations and partnerships; and documentation for policies, procedures, and standards. Together, these four elements combine to ensure a successful digital preservation program within any institutional framework.

Support and Buy-in from Stakeholders

Nearly all attendees of the Best Practices Exchange agreed that the first step in establishing a successful digital preservation program is gaining support and buy-in from stakeholders including the state legislature, internal management, internal IT staff, content creators, vendors, and end-users. Participants noted that the two most important factors for gaining support are legislative mandates and education.

Throughout all the small-group exchange sessions, attendees emphasized the importance of legislative directives mandating the maintenance and preservation of digital assets. A legislative mandate serves several critical functions. First of all, it has significant influence when garnering support and buy-in from both internal and external stakeholders. Moreover, it provides a tool to enforce compliance among uncooperative stakeholders. Finally, mandating digital preservation paves the way for the establishment of sustainable funding. For many of the attendees, a legislative mandate and sustainable funding were the two key ingredients for a successful digital preservation program.

Even so, a legislative mandate does not guarantee stakeholder cooperation. Education is another way that digital practitioners can gain buy-in from internal and external stakeholders. Gaining buy-in from content creators through education allows for better and more efficient collection development, as well as a greater likelihood of following the legislative mandate. Meanwhile, educating vendors and product developers helps in the creation and maintenance of customized systems. Finally, educating end-users, such as depository libraries, ensures that user needs and expectations are being met. Several Exchange attendees described outreach efforts with state agencies that included advocating the importance of digital preservation, reminding agencies of appropriate legislative mandates (if any), and teaching agencies about methods of digital publication/record submission. As one participant noted, “it is critical that everyone who touches state government information throughout its life be aware of their role in preserving it for the future.”

Education efforts are also needed within the libraries and archives to gain buy-in from management and IT staff. Gaining management support is critical because it allows institutions to unite in their commitment to preserve digital information. It is equally important to gain buy-in from internal IT staff who possess a level of technical expertise not typically held by librarians, archivists, or records managers. Many participants noted that their internal stakeholders were reluctant to support digital preservation because of the associated costs . In fact, one participant described how his management viewed digital efforts as a “black hole sucking in endless amounts of money.” To combat this misperception, the participant educated his management staff on the cost of not preserving digital information. He described a doomsday scenario in which the state’s data structures were to collapse. Without preservation back-ups, the entire data structure would have to be rebuilt manually, and that would end up costing far more money than ongoing digital preservation and backup. This cost/benefits analysis ultimately led to management support and buy-in.

Figure 2. These images from the murals of Eileen Clegg illustrate the many aspects of gaining support and buy-in from stakeholders.

“Good Enough” Practices Implemented Now

By the end of the Best Practices Exchange, it was abundantly clear that there were no “best practices” for digital preservation. Instead there were merely “good enough” solutions that were being implemented until something better and more effective was discovered. Several participants stressed the importance of “good enough” practices, insisting that practitioners must work with the tools available to them in order to manage and preserve digital information to the best of their ability. Using “good enough” practices does not mean settling for sub-standard systems or results, though. It means implementing the best-possible solutions that are available with the expectation that new resources, solutions, and technologies will be implemented as they evolve. Moreover, “good enough” practices must be flexible, modular, and interoperable among technical and organizational infrastructures.

Flexibility is a critical component of any digital preservation program. Returning to the analogy of the shifting sands in the “digital desert”, it is important that a practice be flexible enough to avoid being buried under the constantly changing sands. Similarly, any “good enough” practices must be scalable so they are capable of growing and evolving to meet future demands of usage and technology.

Participants agreed that “good enough” practices must be modular with various components that can work together or independently. As technologies change and evolve, certain components can be changed or upgraded without requiring an overhaul of the entire system. One participant compared an effective preservation program to a set of Lego building bricks: each brick is a unique practice that can be attached to others, changed, and/or removed to suit the current technologies and user needs.

Finally, participants stressed the importance of devising interoperable solutions, systems, and standards to allow for easier collaborations and joint projects in the future. Interoperability is critical to make information readily available and accessible across institutions. Interoperability is most successful when built into standards, solutions, and systems early on in the planning process. One participant noted that during the development of a digital preservation program, it is necessary to think globally when acting locally.

But why try to tackle digital preservation with “good enough” solutions? Why not wait for the perfect solution to come along? All contributors agreed that waiting for the perfect solution was neither feasible nor desirable. The digital history of government agencies from the 1990s is gone forever and more digital information is disappearing every day. Waiting for “the perfect solution” will only result in the loss of more digital information. One practitioner noted that perfect solutions only exist in a perfect world, so practitioners must use what is available now. In other words: don’t let the perfect prevent the possible.

Collaboration and Partnerships

Digital management and preservation require more resources than any single institution can sustain. As such, successful digital preservation programs need to include partnerships within and across institutions. Exchange attendees described collaboration by using adjectives such as “essential,” “key,” and “critical.” Collaboration allows institutions to share responsibilities, resources and, most importantly, results.

Digital preservation programs are most effective when they are built with partnerships in mind. Using interoperable systems and solutions will greatly enhance the likelihood of collaborations with stakeholders such as content creators, state agencies, depository libraries, vendors, and other institutions. Partnerships may be short-term, only lasting the duration of a specific project, or may be indefinite depending on the overall goal of the collaboration. Either way, all collaborations should be clearly documented with well-defined roles, expectations, and results. Maintaining this documentation will decrease the likelihood of future problems or misunderstandings cropping up.

Inevitably, though, disagreements will occur between partnered institutions. Participants listed the most likely causes of disagreements as: confusion about the role of the partnered institutions; misunderstandings in terminology; systems incompatibilities; different standards; and a lack of social and technological interoperability. Fortunately, the benefits of collaboration far outweigh these difficulties. Collaborations offer fresh perspectives and ideas that enable digital practitioners to think outside of the political and cultural mindsets of their institutions. These fresh perspectives lead to richer solutions and better outcomes. 
 
But how does one go about identifying potential collaborators? Attending national meetings (such as the Best Practices Exchange) and workshops (such as Cornell’s Digital Preservation Management Workshop) is one of the most effective methods for building partnerships and fostering collaborations. Meetings and workshops give practitioners the opportunity to learn about new projects, programs, systems, and solutions while identifying institutions with common goals and interests. Indeed, practitioners noted that frequent meetings and workshops help in the establishment of a “community of practice.”

One participant characterized a “community of practice” as a flock of birds. Each bird may ultimately have a different end destination, but since they are flying in the same general direction, it is more efficient to fly together as a flock. Like a flock of birds, working together as a community of practice allows for shared knowledge and resources, a more efficient route, and a faster arrival time. But a community of practice goes beyond attending the occasional meeting or workshop. It must be a continued effort to work together with other institutions to share knowledge and ideas. Adam Jansen of the Washington State Archives believes this is best accomplished through the use of a robust website. During the Exchange, he debuted his idea for a Digital Preservation Network SharePoint [2] site where practitioners can share information about their projects, learn from other institutions, and identify potential collaborators for the future.

Figure 3. These images from the murals of Eileen Clegg illustrate the many facets of collaboration.

Establishing Policies, Standards, and Procedures

Documenting policies, standards, and procedures is the fourth and final element for a strong foundation in a digital preservation program. Participants felt that policies should clearly document the value of digital preservation within the institution while addressing born-digital, digitized, and analog formats. A well-written policy should serve as historical proof of an institution’s commitment to digital preservation now and long into the future. Moreover, the development of policies, standards, and procedures should be done in conjunction with the other three elements of the foundation of a digital preservation program. Documentation should be developed collaboratively with stakeholder buy-in to ensure support while using the “good enough” practices currently available.

But how do you go about developing policies? Several Exchange attendees noted that the process should begin with the creation of a policy framework including an overview of how digital preservation will be integrated into the institution’s goals and objectives. Once management staff approves the framework, it will serve as a blue print for the development of more specific policies and guidelines. All policies should reflect the vision presented in the framework.

After policies are developed and implemented, it is important to address standards. All aspects of digital management and preservation are affected by standards including metadata schema, file formats, submission methods, and copyright issues. It is important for an institution to carefully research all options before establishing standards. However, once standards and best practices are determined, it is essential to document them in a clear and detailed manner. These standards will be distributed to stakeholders outside the institution, so it is important to define all terminology to avoid semantic misunderstandings.

Finally, it is important to plan and document procedures that describe each process related to the management and preservation of digital information. These processes should be simple yet flexible, so that there is a diminished learning curve and minimal impact on staff and resources. Digital practitioners should consider adapting current practices used for analog material to fit with digital formats as an alternative to “reinventing the wheel.” However, if current practices do not fit with the needs of the institution and its stakeholders, digital practitioners should be willing to look beyond current practices in order to develop the best procedures possible. Naturally, developing and planning procedures takes time. Not all institutions have this luxury, so they may need to devise quick and dirty procedures for immediate use until they can afford the time and resources to plan more efficient processes.

Once policies, standards, and procedures are established for all aspects of digital management, it is critical to share these documents with partners, as well as internal and external stakeholders. Disseminating this information allows for greater buy-in, support, and compliance.

Conclusion

The Best Practices Exchange 2006 offered attendees a truly participatory environment to share knowledge, ideas, project updates, solutions, and lessons learned in the pursuit of digital preservation. The Exchange emphasized the concept of the “community of practice” and the need to communicate and collaborate across institutions for better results. This sense of community was showcased throughout the Exchange when attendees described the requirements of a successful digital preservation program. Regardless of their level of knowledge, their experience, or their institution, they all voiced a similar opinion: a strong foundation is necessary for a strong digital preservation program.

This consensus demonstrates that digital practitioners have only just begun to tackle the monumental problem that is digital preservation. Members of this community of practice recognize that they are the architects responsible for building the future of digital preservation, and they understand the importance of laying a stable groundwork for long-lasting results.

The success of the Best Practices Exchange has prompted several institutions to ask about hosting the Exchange in the future. This interest bodes well for digital practitioners, suggesting that the Best Practices Exchange will become an annual function and that the community of practice will continue to grow and evolve as they jointly seek solutions to the dilemmas of digital preservation.

Notes:
[1] To learn more about the Best Practices Exchange 2006, please visit the website: http://statelibrary.dcr.state.nc.us/digidocs/bestpractices.
[2] The Digital Preservation Network SharePoint site is due to debut by the end of June and will be located at the URL:  http://www.preservationnetwork.org.

UBdigit is a developing local repository—an asset management infrastructure—for a variety of University at Buffalo (UB) digital material including course content, image collections and archival holdings. Like other universities, UB is witnessing an evolution of digital collection building from highly idiosyncratic digital projects to controlled environments for the development of select unique content. UBdigit is facilitating campus-wide migration from isolated digital projects to integrated digital collection development within an nascent institutional repository environment.

Campus Context

The campus is becoming an increasingly media-rich interdisciplinary learning space that mandates ready access to not only digital texts but also visual and audio materials. Faculty and learners wish to choose among an ever broadening array of traditional research and teaching materials, including digital media objects. The campus teaching/learning culture is marked by a new media reality in which we see the ascendancy of media learning objects across disciplines and where “reading” media is a legitimate scholarly activity. Students entering the university are representative of an arguably more media-literate millennial generation who view digital content as integral elements of their academic work. Digital content is being created in a variety of formats for delivery to users over networks, on removable storage media, and on a variety of mobile computing devices. Dependable and easy access to these materials in part provides both the backdrop and the motivation for university-wide collaborative development of a digital asset management project such as UBdigit.

Growing demand for digital content has resulted in the proliferation of idiosyncratic local digitization projects. Historically, these digital projects have grown out of faculty scholarly interests or library special collections and have been produced to fulfill a specific local purpose, such as teaching, research, access, and/or preservation. Among them are many highly-valued digital collections that facilitate shared content but often they lack content management, are wasteful of institutional resources for digital efforts, and are not now scalable or compliant with emerging repository standards. Additionally, these collections have not been developed or optimized for integration with larger projects, other digital collections, or repositories.

By applying asset, data, and infrastructure standards to existing and new digital collections, UBdigit provides a controlled environment that facilitates and promotes integration of our project-based digital content with larger digital library and repository content.

Figure 1. Adopting standards facilitates asset integration.

Differentiated Needs and Supports

Faculty, librarians, and IT professionals share an interest in the management of digital content – particularly that which is intended for sharing with end-users and for integrating with other information systems. However, we also recognize that not all digital content requires, can, or should receive the same level of treatment (e.g., development, management, distribution, and commitment to preservation and persistence). We differentiate between those assets for which we will provide responsible digital asset management for access, discovery, and use and those for which we will assume institutional repository commitment to persistence.

In either case, agreement on standards for assets and associated metadata records, coordinated development of digital collections, provision of expert services, and provision of scalable organizational and structural environments to sustain collections over time are needed. We have implemented an integrated approach to provide differentiated support to UB digital collections:

UBdigit contains “institutional repository” content—those assets for which we assume a long-term, preservation commitment. These collections contain unique objects for which we are the  “institution of record.” UBdigit provides archival quality master digital images and high-level metatdata for these unique collections to promote broadest discovery and use.

The “UBdigit Teaches” segment contains content primarily used for a teaching purpose. It is a utility-driven service to present digital surrogates of analog content previously held in isolated departmental or faculty collections. The goal of UBdigit Teaches is to make these materials more accessible to all university constituents within a secure institutional network. Contributions to UBdigit Teaches are selectively developed against holdings in other university-owned or licensed proprietary collections. For most of these objects we are not the institution of record, and, therefore, we offer much more limited metadata development–sufficient to facilitate discovery and access. Digitization of these materials is toward optimum access and re-use within a variety of presentation and distribution environments.

Development of UBdigit

UBdigit grew out of a faculty-driven community mandate for an infrastructure for shared collections. It is based on a scalable economic principle of distributed responsibility and shared benefits with three main conceptual goals:

1. Provide a robust, persistent campus-wide repository for UB’s archival, research, and teaching collections in digital form.
2. Provide responsible stewardship for the creation, description, management, storage, and distribution of digital assets.
3. Emphasize utility, standards, access, and economy of scale.

The institution augmented these conceptual goals with administrative mandates to promote efficient and economic use of institutional resources (investments, time and effort), avoid redundancies and focus on unique contributions, utilize broad base of institutional expertise and capacity, and optimize access to content.

In an effort to meet the expectations of our community and our administration, we conducted an institutional scan to analyze local institutional culture, to identify shared goals, and to establish an inclusive foundational network of partners. These include:

• Champions – providing institutional support and guidance for the process
• Visionaries – providing phased planning and project management throughout the planning lifecycle
• Contributors – providing content and expertise throughout project implementation

UBdigit was envisioned, developed, and brought to production in four phases. It is as much an infrastructure for shared collection development as a strategy for introducing a standards-based approach to collection and item level project development.

Phase 1: Development
Three discovery sub-groups (including faculty, visual resource staff, Information Technologists, and library service providers) were established: 1) technical infrastructure, 2) standards for assets and metadata, and 3) project policies and protocols. Each discovery group produced formal reports, which subsequently became project workflows and policy statements. Utilizing a common features/functions checklist and an independent review process, the infrastructure discovery group made a consensus recommendation of CONTENTdm [1] for our underlying data infrastructure. We tested the application on a small data and asset set using the OCLC-hosted environment for both our alpha and beta testing cycles.

Phase 2: Operationalization
Discovery groups prepared standards documentation for digitization, metadata schema, a comprehensive data dictionary, and a metadata crosswalk. These documents are continually amended to accommodate new formats, standards, object types, and metadata schemas as they are implemented within UBdigit.

In phase two, we also developed graphical models and process workflows, including our repository model (see Figure 2), user access model, proposal and review process workflows, etc.

Figure 2. UBdigit access and integration model.

Phase 3: Implementation
The University Libraries host UBdigit on redundant sandbox and production server environments. Library staff provide server and application support including application customization, GUI development, documentation, beta and production rollouts and data loads, quality assurance checks, and application version migrations.

Phase 4: Production
We currently host two repository segments: UBdigit (public access repository) and UBdigit Teaches (authenticated access to UB community only). Collections are drawn from the UB Libraries, institutional archives and special collections, and from academic department and faculty collections. Content currently includes still images and will soon incorporate audio and time-based media. In the future we anticipate kinetic images, animations, virtual-reality, interactive sequences, and multimedia constructs—in essence, any Web-deliverable media content that serves a teaching or scholarly purpose.

Both UBdigit and UBdigit Teaches collections are selected and prioritized for development from contributing units and individual faculty across the campus. Each proposed collection undergoes a pre-production evaluation process to determine development parameters, specifically scalable collection-level standards paired to institutional commitment that set realistic minimums/maximums for economic production and file maintenance over time. Some evaluation criteria include:

• Degree of access - authorization/authentication requirements based on intellectual property conditions
• Depth of metadata - development needed to facilitate discovery and to enable integration with other digital collections to the appropriate degree versus “institution of record” requirement for full metadata development
• Digitization targets - for archival master files and distributable services files
• Commitment to persistence - e.g., “trusted digital repository” commitment, or network redundancy, or highly available/offline/near line access to redundant copy

The Libraries are the UBdigit stewards and provide central support for the persistent environment. The Libraries are responsible for UBdigit governance, advocacy for resource allocation, project management, standards development and quality assurance, systems and applications support, network capacity, security, and archiving.

Contributing units provide content and support for UBdigit from libraries, visual resource collections, academic collections, and museum collections. Contributors assume responsibility for digital collection curation, metadata development, and content authority for individual collections within UBdigit.

To see UBdigit in production, please visit our website.

Figure 3. UBdigit website screenshot.

The “About” section of the website provides documentation on each phase, as well as contact information and policy documents. You can browse UBdigit by collection or through a variety of search interfaces from keyword to advanced with delimiting functions throughout. We currently host a variety of collections developed to support the teaching, archival and research interests of the university and to provide visibility for unique institutional holdings. Some examples are detailed below.

Evolutionary Biology Digital Dissection Collection
Co-developed with the Department of Biological Sciences

This collection captures born-digital dissection imagery developed in 1998 for use and distribution through a stand-alone course Web page. The digital content was repurposed in 2000 for a mastery-learning component within a course management system self-assessment module. In UBdigit, the images were disaggregated from the linear CMS presentation. Each multi-view animal dissection is presented as part of a compound object set; each view can also be searched and viewed independent of the compound set, allowing users to create combinations and learning sets to meet their own learning needs.

Figure 4. Compound object menu screen.

Pulp Fiction Cover Art Collection [2]
Co-developed with the Arts & Sciences Libraries and the University Libraries Special Collections

The UB Libraries’ George Kelley Paperback and Pulp Fiction Collection is one of the largest (over 30,000 volumes) collections of well-preserved pulp fiction in the world. This UBdigit collection provides researchers with JPEG2000 (pan and zoom) views of the cover, back, and spines of this graphically-intense, mid-twentieth century, mass media art form.

Figure 5. Sample cover art: Beware the Lady, Cornell Woolrich (author),
George Kelley Paperback and Pulp Fiction Collection.

Log of the Good Yacht Althea on the Summer Cruise of 1914
Developed by John Bewley, UB Music Library

One-of-a-kind legacy objects are preserved through digitization and presented for wide access. This typescript sailing log was written by Louis Wright Simpson and consists of 46 typescript leaves with thirty 2“x 3" photographs attached. This UBdigit OCR-scanned, full text resource allows users to explore the log page by page, or by keyword and delimited searching functions.

Figure 6. Digital preservation OCR scan of archival text.

Edgar R. McGuire Historical Medical Instrument Collections
Co-developed with the UB Health Sciences Library History of Medicine Collection

The collection contains more than 150 instruments (primarily 19th century) or sets of instruments chosen for their illustration of past medical and dental procedures.

Figure 7. 19th century syringe, Joseph Frederic Benoit (designer),
McGuire Medical Instrument Collection.

UBdigit Next Steps

UBdigit and UBdigit Teaches are actively developing new collections using an interdisciplinary contributor model. By inviting collaboration among variously expert partners, we are able to grow both in volume and diversity. Among our near-term challenges are achieving interoperability with other UB and proprietary digital asset collections and experiments with data harvesting, extraction, and ingest. We hope to achieve a higher degree of integration with external digital library and repository initiatives. Within UB and SUNY there is growing interest in both shared collection building and institutional repositories. Because UBdigit adheres to rigorous metadata standards and leverages OAI harvesting protocols, it is useful both as an infrastructure model and as ground for integration with experimental initiatives, such as SUNY DSpace, [3] and to explore broader integrations with collectives, such as ARTstor and the National Science Digital Libraries.

For more information on UBdigit,visit our website or contact us at ubdigit@buffalo.edu.

Notes:
[1] OCLC provides licensing for CONTENTdm digital collection management software.
[2] Most of the cover art in this collection corresponds to texts analyzed in the UB Libraries’ Gumshoes, Sleuths & Snoopers database at: http://libweb.lib.buffalo.edu/kelley/KelleyAdvanced.asp. For more information on the George Kelley Paperback and Pulp Fiction Collection, see the website at: http://ublib.buffalo.edu/libraries/units/lml/kelley/.
[3] SUNY DSpace, digital repository of the State University of New York, see http://dspace.sunyconnect.suny.edu/.

With several years of experience making images available for academic and research use, and conducting user studies and focus groups, RLG has learned quite a bit about digital image services.

In the late 1990s, RLG began investigating ways to work with members to make their special collections more accessible to RLG members and new audiences. These investigations led to discussions with the RLG board and members that resulted in a decision to form the Cultural Materials Alliance, a subset of RLG members interested in pursuing ways to make their special collections available in digital form.

Figure 1. Screenshot of the Cultural Materials website.

The Alliance took shape in 2000 and formed advisory groups to collectively address topics such as policy issues, descriptive recommendations, digitization recommendations, and collection development. The first collections were assembled; a brand new platform and an interface was developed; and the Cultural Materials service was launched in 2001.

In a parallel effort, RLG hosted the AMICO (the Art Museum Image Consortium) testbed, beginning in 1998. AMICO launched the testbed to study the utility of digital images in the academic environment. This successful pilot led to the AMICO Library, which RLG made available for the next six years until the AMICO consortium disbanded. At that point (July 2005), RLG released CAMIO (Catalog of Art Museum Images Online) in order to continue to fill the need demonstrated by the enthusiasm of AMICO users.

Discovery and Use of Images

When the Cultural Materials service was developed, we took special care to create an interface that would support discovery of these materials. We created several ways to browse the whole resource and to get an overview of a result set. We created thumbnail displays, offered relevance ranking and other ways to sort results, and devised ways to view high-resolution images and multi-part items. We had input from a member advisory group and a panel of technology experts. We also had an independent analysis conducted of usability issues. In the end, we were quite pleased with the interface and functionality—and got rave reviews from many information professionals. But there was one problem: Cultural Materials service was getting relatively low use.

Figure 2. Screenshot of the Cultural Materials website.

Faculty, especially in the arts and humanities, are keen to use images in the classroom. Many have used slide collections for years and are eager adopt the advantages offered by digital images. At first technology was a barrier; many faculty did not have access to wired classrooms and others did not trust the network reliability. Now, however, those concerns are voiced less frequently. When asked about image use, faculty report most commonly that they use their own photographs and scanned images, as well as images found on the Web. When asked about licensed resources, many do not even consider that their library might license image resources on their behalf and wouldn't know how or where to look for them if they did. When informed about Cultural Materials, CAMIO, and other resources, they complain that they can't look in multiple places, especially when each resource has only a few relevant images. Most report they could often find something adequate by searching with Google.

If faculty members aren’t using Cultural Materials, they probably aren’t suggesting that students use it either. The library has too many e-resources to effectively promote them all to likely users. The obvious solution was to make the images available where users were looking, and the most effective way to do this was to make them available to Web indexes, such as Google and Yahoo.

From the beginning, the Alliance expressed a desire to first serve the academic research community, but then to reach out to new audiences. We could do both of these things by making a version of the content discoverable on the open Web. Trove.net was developed to increase the use of Cultural Materials content. We excerpted some metadata that could be crawled and indexed. (A few of the Cultural Materials collections have rights or licensing situations that precluded their inclusion in Trove.net, but the vast majority of Cultural Materials content is in Trove.net.) Trove.net displays a watermarked, “snapshot” size image (250 pixels in the longest dimension), along with descriptive metadata and links into Cultural Materials for subscribers or links to licensing information for others. Now, people looking for images on the Web can discover Cultural Materials content–and discover the richness of the service.

Figure 3. Trove.net website.

Figure 4. Trove.net item.

Another way to bring the images to places where users are looking for them was to acknowledge that many campuses have invested heavily in image solutions, such as ARTstor, MDID (the James Madison Digital Image Database), Luna Insight, or an image management system. We have begun investigating ways to make Cultural Materials discoverable from these other systems. Experiments are now underway with ARTstor and with MDID.

Despite the acclaim our carefully constructed interface received, we ended up doing everything possible to help users find Cultural Materials items without it!

We also learned that a primary academic use of images is in classroom presentations. While this usually requires presentation software, we didn’t think it made sense for us to create such software. Instead, we sought to make it easy to use our images within presentation software already in use. The work underway to make Cultural Materials interoperable with ARTstor and MDID will address this for those systems. But because many users use Microsoft PowerPoint, we made it easy to export high-resolution images for use in PowerPoint and created an email feature that enables the user to send the image, a caption, and a CSV file of slightly more detailed caption data. (We also provide a download button for the highest resolution images, in case email systems cannot handle the file size.)

Figure 5. High resolution image view with email option.

Thus, the new focus is on discovery of Cultural Materials images in places users normally look and allowing Cultural Materials images to be used with the tools users normally use.

Getting the Word Out

Having addressed our most important user issues, we then wondered how we could make Cultural Materials available to other audiences. RLG members primarily purchased institutional subscriptions to Cultural Materials—most were large universities, many of which had also contributed to the resource. CAMIO subscriptions, however, were purchased by a larger variety of institutions, including those that had never bought RLG resources before. Why were we having more trouble selling Cultural Materials?

CAMIO had an audience with a need. Art History teachers were teaching a highly visual subject matter. They had long used analog slide presentations and were early adopters of digital images, relying on visual resources curators to support them. Art libraries had budgets to support the acquisition and use of visual materials. We knew who the interested parties were and how to communicate with them.

Who needs Cultural Materials? Cultural Materials has many subject strengths: History of Science (Medicine, Aviation, and Astronomy, for instance), Architecture, Anthropology, Maps, History of Photography, Graphic Arts, Labor issues, Performing Arts, and so on. But who needs all of that–or rather, who would lobby to license such a wide-reaching resource? Its unfocused, multifaceted nature made it difficult to describe and less compelling as a solution to a familiar need. And yet it’s because of the synergy across subjects that Alliance participants believed Cultural Materials would be especially valuable.

Because Cultural Materials was serving as a vital stage for addressing issues of member concern, RLG was willing to heavily subsidize its development. Now that it is a mature resource (and very expensive to maintain), it needs to pull its own weight. New efforts will be launched to reach additional institutions by targeting content and marketing to the needs of specific communities. User focus groups and surveys will guide these efforts.

Finding New Markets

Along with reaching new audiences, Trove.net seeks to address another goal of the Alliance members: to explore new sources of revenue. If successful, revenue realized from the use of specific images will be shared with the contributors of those images. General revenue will subsidize the subscription service.

Many contributors already had licensing and reproduction services, but use and revenue was low because potential licensors had to know the image existed and know where to find it. Some contributors were overwhelmed by requests and were happy to have RLG take the load from them. Others were satisfied with their own service, but welcomed new ways to reach potential licensors. And yet others hadn’t even dipped a toe in these waters and were glad to have the opportunity to do so.

RLG considered developing a stockhouse service internally, but realized that the most significant demands would be in marketing to an audience we didn’t know and who didn’t know us. By partnering with established stockhouses, we would have experienced service providers who already had a loyal client-base.

Trove.net images would link to the stockhouses set up to manage the licensing, but more importantly, Trove.net images would be in the stockhouses’ systems where commercial clients already go to find images (a familiar theme: put the images in places users normally look). Deals with multiple stockhouses would reach more clients. Our first stockhouse partner was IndexStock, a sort of mid-tier stockhouse that was willing to accept all of our content. This gave all of our contributors a chance to learn about the licensability of their images. (Top-tier stockhouses are highly selective, likely to choose fewer than 5% of our images based on their assessment of which images would be licensed.) Another advantage to IndexStock is that they have a service called Photos To Go that licenses lower resolution images for limited uses for a much lower cost. This was especially important because many Cultural Materials collections were scanned years ago and were captured at fairly low resolutions. PhotosToGo could keep those contributions in the game.

Figure 6. IndexStock website.

RLG was also interested in other revenue possibilities. We needed to start covering the cost of running the Cultural Materials service without raising subscription prices. Was there potential for selling advertising on the Trove.net site? Were there Internet providers who would license content? Were there other potential partnerships to investigate? We employed a number of consultants to investigate other stockhouses, search engine companies, content syndication sites, research and reference sites, publishers, and other types of e-commerce possibilities.

To date, the most promising of these have been the publishers–not all that distant from stockhouse licensing, which was where we started. The heyday era of sites willing to pay for content has given way to sites accepting paid inclusion. We had to go out of our way to get indexed by search engines and included in image search sites. An experiment with putting Google AdSense ads on the Trove.net site has had only modest returns. An arrangement with LookSmart to include Trove.net images in FindArticles results for a percentage of their ad revenue has had similarly modest results.

In addition to forging additional stockhouse arrangements, future efforts will target increasing traffic to the site. To that end, we’re experimenting with Google AdWords and we are doing what we can to raise our visibility and ranking in Google. Future partnerships will most likely focus on increasing traffic.

Increased traffic will boost ad revenue and licensing revenue. If we can get sufficient traffic levels, we’ll implement PayPal transactions for easy licensing of images for personal use. Conceivably, significantly increased traffic may interest corporate sponsors. And increased traffic will drive more users to the Cultural Materials service.

Cultural Materials and Trove.net do not exist to make money, but to serve the academic community. We hope these attempts to maximize alternate revenue sources will result in a new business model that will reduce subscription fees, while enabling RLG to fulfill its mission of improving access to the rich research collections held by member institutions.

UPDATE:  This article was written during RLG’s planning for its next fiscal year. Since then, RLG has combined with OCLC (see the press release). Currently, staff of both organizations are discussing the future of CAMIO, Cultural Materials, and Trove.net. The services will continue to be offered. Contributions will be sought from a broader array of libraries, archives, and museums. RLG and OCLC staff will consider how to leverage some of OCLC’s other digital collection services, including CONTENTdm, Digital Archive, and the Content Cooperative Pilot to enrich the experience of institutions that subscribe or contribute to Cultural Materials. Together we will explore additional ways to reach broader audiences via Trove.net.

Greenstone is a suite of software for building and managing digital library collections and for distributing them on the Web or on CD-ROM. Like other important digital library tools, such as DSpace and Fedora, Greenstone is an open source development, meaning that the software is designed and written by a community of developers; the software is frequently updated with improvements and fixes; the source code is freely accessible, making possible extensive customization with plugins; and the software is available to download and use at no cost under the GNU License. Among Greenstone’s strengths are that it has fairly low technical requirements and good supporting documentation, it runs on multiple platforms, it is available in multiple languages, and it offers support for a wide array of metadata options.

Greenstone 2.7 was recently released, as well as a new and extensive documentation wiki. (Greenstone version 3.0—a complete redesign, written in Java, and with better integration with XML-based technologies—is currently being tested.)

Greenstone’s  website features: links to download the software, links to several example implementations, an FAQ, as well as full documentation and other support devices.

Greenstone is produced by the New Zealand Digital Library Project at the University of Waikato, and it is developed and distributed in cooperation with UNESCO and the Human Info NGO.

It has been over 10 years since the federal government developed policies governing management of email records by its agencies. What is the current status of policy development in the 50 states?

Note: This is Part II of a two-part FAQ. Part I covered the US Federal government's laws and regulations. This part II surveys regulations and policies in the 50 United States.

Introduction

In Part I of this FAQ, we examined the impact of federal email management regulations on politicians and corporations. We saw that poor compliance with existing regulations can have serious consequences, but also how email retention policy can present a major "damned if you do, damned if you don't" dilemma. Although there are similar concerns at the state level, in this part we shift focus to policy development practices and the policy development environment.

The relationship between federal and state laws can take a number of different forms. States that operate programs with funding from Washington may be subject to federal regulations. When federal investigations target state officials (elected or otherwise), federal criminal statutes will apply. However, even in areas where there isn't direct control, there may still be influence. Sometimes federal laws or regulations set a standard, which is then followed, to one degree or another, by the states.

In the case of email management, the background information provided in Part I on the development of federal policy has considerable relevance to the adoption of state-level policies. Prior to Armstrong v. Executive Office of the President, the landmark federal case decided in 1993 that led to new regulations from NARA (US National Archives and Records Administration) governing retention and management of email messages by federal agencies, there had been substantial disagreement over whether email messages qualified for record status. In fact, three successive US administrations (Reagan, Bush Sr., and Clinton) had argued during Armstrong that email messages were not records. This may seem like an absurd contention today, given the critical role that email records have since played in criminal prosecutions and freedom of information requests, but, in the early 1990s, the degree and sophistication of email use in government was highly variable. Even though many archivists and records managers recognized that email messages could qualify as records, protocols and systems for handling record email were not well established. Early email systems were often isolated from other aspects of the information technology infrastructure. (An article from a 1995 issue of Government Technology magazine entitled "Should Agencies Archive E-Mail?" illustrates the confused picture at the state level in the immediate post-Armstrong period.)

Ultimately, records managers at the state level took their cues from NARA. When NARA clearly defined email messages as potential records, many states started developing their own policies. Ten of the policy documents profiled here specifically reference Armstrong, and none of the state-level email management policies reviewed here appears to have been adopted prior to 1995.

However, even though states have looked to both federal policy and each other's policies for guidance in developing their approach to email (see sidebar), there is considerable variation in the current status of policy development in the 50 states. More than 10 years post-Armstrong, with much having been written about the best archival and records management practices for email, there are still no truly uniform standards for the retention and management of email messages among the states. In this FAQ, we take a closer look at the range of policies in place.

Why Focus on Email Policies?

Given the bedrock tenet in records management that function, not form, determines whether an information object qualifies as a record, and for how long and in what manner it should be retained, a focus on email may seem misguided. Many states rightly argue that email is not in itself a record series, but rather a means of communication that can produce records (as well as non-records) of many different kinds. But email, like all electronic records, has unique characteristics and poses unique management challenges that call out to be addressed.

1) In most organizations today, email is sent and received in huge volume, with content that includes unwanted spam, informal banter (with family members, friends, and colleagues), scheduling and documentation of meetings, transactional and contractual correspondence, and official policy statements. In most email systems, all types of messages are intermingled.

2) Email is both easier and harder to destroy than paper documents. It's easier in the sense that large numbers of messages, including those that qualify as public records with historical, documentary, and legal value, can be deleted in seconds. It's harder in the sense that messages that should be shielded from being swept up in public records requests, such as personal correspondence, may exist in so many backup systems that they are impossible for the sender or recipient to completely eliminate.

3) The means for maintenance of context for email records may not be obvious or straightforward. For example, address book entries may employ aliases for large groups of recipients whose real names do not appear on the message itself. Blind copying may obscure the existence of some recipients. Attachments and embedded multimedia content may not automatically accompany the relevant message.

4) Despite the presence of standards for the structuring and encoding of email messages, incompatibilities based on computing platform, software, and configuration can cause loss of data and misinterpretation of characters.

5) Storage of email in a form that can be truly responsive to public records requests, the documentary needs of staff, legislators, executives, and, ultimately, historians, requires the use of specialized records management systems. Such systems can be costly and onerous to establish, especially when there are large numbers of legacy email records.

Thus, policies for the proper management of email are most relevant when they address its particular machine-readable qualities and not just apply guidance developed for the equivalent type of paper record.

Policy Development Environment and Rationale

One of the challenges in developing policies for email management is the range of underlying provisions that must be incorporated and the number of agencies and other units of government that need to be involved. Existing statutes and case law related to records retention, public records access, freedom of information, and records management will form the legal basis. Regulations addressing access, acceptable use, security, privacy, and confidentiality of email systems will also come into play. Policy statements, handbooks, and training materials have to be developed, and training sessions must be organized and taught. The technical infrastructure has to provide the tools to support the policy. Members of the executive and legislative branches may need to be consulted, along with staff from records management, information technology, and human resources. Not surprisingly then, a look at individual policies in the 50 states will find that they are, quite literally, all over the map.

Adding to the complexity and the diversity of approaches is the fact that state governments are organized in different ways, assigning responsibility for records management to a broad array of divisions and departments. In some states, records management is a function of, or a division of, the state library and/or archives, which may be a high level agency (Alabama, Arizona, Connecticut, Kentucky, Maryland—permanent records only, Mississippi, Oklahoma, South Carolina, Texas, Virginia) or a unit within the secretary or department of state's office (Delaware, Florida, Georgia, Illinois, Louisiana, Maine, Massachusetts, Missouri, Montana, Nebraska, New Hampshire, New Jersey, Oregon, Rhode Island, Vermont, Washington), the department of education (Alaska, New York), the department of cultural affairs or resources (Nevada, North Carolina, Wyoming), or the department of administrative services (Utah). Alternatively, records management may be separate and distinct from libraries and archives and may appear under a public records commission (Indiana, New Mexico), general services (California, Hawaii, Tennessee), administration (Colorado, Idaho, South Dakota, West Virginia, Wisconsin), or information technology (North Dakota). In still other cases, records management responsibility is assigned to the state historical society, commission, or center (Arkansas, Iowa, Kansas, Michigan, Minnesota, Ohio, Pennsylvania), which may be either a part of state government or a private, non-profit organization that has statutory authority to manage the state's records.

Policies that have an impact on email management can emerge at different levels of government and for a variety of reasons. Some states have been working on statewide email software consolidation, an initiative likely to come from an information technology unit, and often motivated by budgetary concerns, but that should also speak to records management needs. Policy development on Internet acceptable use, privacy, and security may also intersect with and require coordination with records management. Conversely, email policies emerging from archives or records management units cannot be done in isolation and should be integrated with information technology policies on software selection and use, and human resources policies on new employee training and continuing education. Collaboration is necessary in order to craft policies that are practical and enforceable from a legal, technological, and staff training/cooperation perspective.

Characteristics of Email Policy Guidelines

Guidance in the development of sound email management policies has been available for many years. One early and quite thorough analysis by David Wallace (see Recordkeeping and Electronic Mail Policy: The State of Thought and the State of Practice), written in 1998, retains its relevance today. Wallace's analysis focused on the following dimensions:

• Record status: Does the policy define what a record is, address the record status of email, and provide criteria and examples to help distinguish record email from non-record email?

• Filing and Maintaining Records: Does the policy specify what means (print, microform, electronic) can or should be used to store record email? Does it offer filing suggestions? Does it address how to handle attachments or encryption? Is there a discussion of long-term preservation and archiving?

• Administration: Does the policy include a link to the records retention schedule? Does it identify the entity responsible for producing the policy? Does it specify sanctions for non-compliance? Is there a glossary of terms used in the policy?

Wallace's criteria provide a good framework for reviewing email policies, but some additional points could be added and some deserve more elaboration.

• Scope: Does the policy identify which branches of state government it applies to or, even better, those that it does not apply to? (Many policies are only applicable to state agencies and not to the executive or judicial branches.)

• Date: Does the policy indicate its effective date and whether or not it supersedes a previous policy? If old policies remain posted, are they identified as obsolete?

• Linkage: In addition to retention schedules, does the policy link to all other policy statements that reflect on email management, regardless of what department or agency produced them? Has it been checked against those other policies for consistency of content and point of view? Are the links to other websites kept up to date?

• Legal framework: Does the policy identify not only which entity is responsible for the document and from what statutes their authority comes, but the entire statutory framework for records management and public access to records?

• Retention vs destruction: Does the policy balance the need for retention of legitimate records with a need to discard non-records? (This avoids both burdening records management systems with superfluous content—spam in particular—and reduces the likelihood that non-record personal messages will be needlessly exposed as public records.)

• Analog vs electronic filing: Does the policy permit filing of record email in an analog form, such as hard copy printouts or microform?

Wallace notes that most policies he reviewed permitted retention of messages as either a paper printout or in electronic form, but that most institutions did not, at the time, have suitable record keeping systems for managing email messages electronically. He also mentioned aspects of some policies that emphasized the importance of including "transmission data" (essentially email metadata such as message date, sender ID, and recipient(s) ID(s)) in any printouts made. Modern email systems routinely include message headers in printouts, but there are a number of reasons to question the use of paper printouts as acceptable surrogates for electronic storage. The inadequacy of paper storage of email messages was noted long ago, and arguments against it have continued to grow in strength:

• Authenticity: Most email clients allow the appearance of messages to be altered. For example, the number of headers visible may be configurable. Furthermore, it is a relatively simple matter to produce an email message that appears to be something it is not. The relatively recent phenomenon known as phishing has taken this process to extremes. Figure 1 compares an email message received by the author as it would be recorded in a printout (shown first) with the actual machine-readable content (shown second). [Note: full IP addresses and domain names have been redacted to protect the innocent and the guilty]. The printout misrepresents (1) the date and time sent, (2) the sender ID, and (3) the content of the message (which employs an embedded image, and hidden "white on white" text, designed to slip the message past spam filters). The printout is completely useless as documentation, yet many state email policies permit, or even prefer, hard copy as the medium for permanent storage of email records. (Another serious threat to the authenticity of email messages, regardless of storage format, is the fact that many email clients allow the stored copy of both incoming and outgoing messages to be modified without creating a record of the alteration.)

• Discrete components: A printout can be easily be separated from external components, such as attachments, embedded images, and nickname files that are needed to form a complete record.

• Storage/Retrieval: The growing volume of record email makes storing it in paper form increasingly impractical. Paper storage also cripples the ability to retrieve messages by multiple search criteria such as sender, recipient, date, and subject(s).

• Content loss: As email becomes a richer medium (e.g., color, embedded multimedia), paper printouts capture fewer and fewer of its characteristics.

Figure 1. Many email policies permit or even recommend printing of messages that are to be retained. This example of a phishing email demonstrates how a printout (shown first) can misrepresent the actual content (shown second) of key elements, including 1) date and time sent, 2) sender ID, and 3) message body.

Conclusion

This two-part series has focused on email management issues related to federal and state law in the US, not because excellent work hasn't been done on email curation in other countries (it has—see the PADI listing in the resources section below), but because email management policy in government is intimately tied to the legal framework for records access. That framework is unique to each jurisdiction.

However, if transparency is a signature characteristic of a government that is accountable, legitimate, and free, then state-of-the-art electronic records management policies and policy implementation are fundamental to good democratic governance, regardless of the specific legal environment. There is no longer any substantial disagreement that email produced by government entities are an important part of the legal and historical documentation produced by governments at all levels.

Nevertheless, despite some very good efforts over the past decade, significant work needs to be done to raise the bar for effective management of email and other forms of electronic communication. An analysis of policies in light of the current electronic communications environment, legal landscape, and implementation strategies reveals that:

• Many policies have not been updated recently and changes in technology have rendered at least portions of them incomplete or out-of-date.
• Training and enforcement tend to lag policy development and are deficiencies that hinder effective implementation.
• Waiting for the law to catch up with technological development does not necessarily produce policies that are the best reflection of archiving and records management principles.
• The tension between records retention laws and the desire of politicians to maintain control over release of their communications is largely unaddressed.

A recent case from a US state is illustrative of several of these points.

Kentucky governor Ernie Fletcher was indicted on May 11, 2006 for criminal conspiracy and other charges in a case alleging that his administration rewarded supporters with state jobs. Email messages are key evidence in the case. In April 2006, as the investigation into the governor's actions proceeded, Fletcher said that he had given up using email for both work and private communication four or five months previously, due to concerns about what would happen to the messages and how they might be interpreted. He was quoted saying "You can't control, you know, what's done on those things. Things can be taken out of context."

One of the governor's spokespeople said that Fletcher was still communicating digitally with his staff using a Blackberry personal assistant that was paid for using campaign funds, but the spokesperson was uncertain whether messages sent or received with it would be subject to the state's public records law. A few weeks later, the state attorney general filed a motion requesting access to the Blackberry messages.

How many states have regulations covering the public records status of Blackberry messages, voice mail messages, or instant messaging logs? A few state policies do, in fact, explicitly mention voice mail (California, Connecticut, Florida, Maine, South Carolina, Texas) and instant messaging (South Carolina, Texas). Some states have separate policies for other forms of communication, e.g. New York has a policy that deals with voice mail alone. Others mention voice mail in their retention schedules. None specifically covers wireless, handheld communications devices like the Blackberry, although some mention wireless communication in their definition of email.

So, despite considerable differences in the freshness and comprehensiveness of their policies, no state should feel too comfortable resting on its laurels. Fortunately, help in development and implementation of policies that reflect current best practices is increasingly available. Just a few months ago (March 2006), the first best practices exchange on digital preservation in state government was held in Wilmington, NC (see story above). The Library of Congress, through its NDIIPP program, in concert with the University of Albany's Center for Technology and Government and IMLS (Institute of Museum and Library Services) has been working with state officials to improve management of digital information assets. NARA is spending $300 million to develop a state-of-the-art Electronic Records Archives, which should have spillover benefits for other government entities. Also, through the NHPRC (National Historical Publications and Records Commission) NARA offers grants geared at improving the management of electronic records for state and local governments. Efforts are ongoing (e.g., Texas, Kansas, and Michigan) to develop electronic records repositories that meet modern operational and functional needs.

We also hope that this review of policy development, along with its resources and listings of state policies, will be useful to state officials and others interested in this issue. The author welcomes any and all feedback, including corrections of inaccuracies and omissions. Please use the email address in the title block for this article.

Click here to download the listing of policies for the 50 states

DCC and DPC Joint Workshop: Policies for Long-term Curation and Preservation
July 3-4, 2006
Oxford, England, UK

The Digital Curation Centre will hold a two-day workshop to explore the range of policies required to manage, preserve, and reuse digital information over time. Concrete examples and real-life experiences of the range and nature of important kinds of policies will be shared through a series of case studies and panel discussions. This event is co-sponsored by the Oxford Internet Institute (OII) and will be conducted at Wolfson College at the University of Oxford.

Data Dictionary for Preservation Metadata: PREMIS Tutorial
July 17-18, 2006
Glasgow, Scotland, UK
 
The Digital Curation Centre will host a two-day, introductory tutorial of the PREMIS preservation metadata data dictionary. The event will cover a walk-through of the PREMIS Data Dictionary and practical implementation issues, including XML rendering. Both days will include hands-on exercises and facilitated discussions to share experiences.

Digital Libraries à la Carte: New Choices for the Future
August 22-25, 2006
Tilburg, The Netherlands

This modular style workshop will focus on four topical areas:

1. Technological developments: Threats and opportunities for libraries
2. Hands-on: Library 2.0 technologies to reach out to the customer
3. Libraries supporting research and Open Access
4. Libraries and teaching and learning

Each module includes four or five lectures by experts in their field. Training formats will include presentations, discussion sessions, and hands-on exercises and activities.

European Conference on Research and Advanced Technology for Digital Libraries (ECDL)
September, 17-22, 2006
Alicante, Spain

The ECDL is commemorating its tenth anniversary with the special topic: “Towards the European Digital Library.” The weeklong conference will feature workshops, tutorials, panel sessions, poster exhibitions, and networking events.

Third International Conference on the Preservation of Digital Objects (iPRES 2006)
October 8-10, 2006
Ithaca, New York

The theme of this year’s iPRES is “Words to Deeds: Collaboration in the Realm of Digital Preservation.” Plenary sessions will explore topics in Preserving Multimedia Objects, e-Journal Preservation, Certification, and National Efforts in Digital Preservation. Concurrent sessions will focus on Tools of the Trade, eScience, Repositories, and New Initiatives. Contributions in concurrent session topics can be submitted by sending a brief abstract to ipres2006@cornell.edu (deadline for paper submissions is August 15, 2006). The deadline for early registration is September 1, 2006.

International Conference on Asian Digital Libraries (ICADL 2006)
November 27-30, 2006
Kyoto, Japan

This will be the ninth International Conference on Asian Digital Libraries (ICADL). ICADL is an annual event that focuses on digital libraries and related technologies. Paper submissions on a wide range of topics, including digital archives and digital preservation, will be accepted until June 30, 2006.