ALA Midwinter discussion (RECON) - January 21,2006
Group 1 Recon – Breakout Session Notes
1/21/06
Ed Shreeves (U Iowa), Dennis Dillon (U Texas), Nancy Davenport (CLIR), Melinda Baumann (U VA), Jeanne Richardson (Arizona State U), Mark Sandler (U Michigan), Laurie Kaplan (CSA)
• The discussion started with Dennis noting that Texas is trying to decide whether to gear up their digitization project, or to shut it down. They are talking to Texas Digital Library and OCA to make the decision.
• Consider a model whereby funds are collected centrally and distributed to those making convincing proposals that meet the goals of the effort – a panel of expert peers vets proposals. Fund governed by group that puts up the capital so that it has legitimacy. The group would not control the process; these funds would be supplemental.
• Some institutions are questioning why they should digitize at all, especially more widely held titles, if Google or OCA will do it. However the quality, accessibility, and longevity of the Google/OCA versions are not clear. The argument can be made that the process should stay in the libraries to be sustainable, broadly usable, available today and tomorrow and a hundred years from now.
• Concern whether who in the library community has the required set of skills to manage this process and get it done
• OhioLink reportedly sent an RFP to vendors requesting digitizing services to be aimed at content identified by OhioLink, with little response. Commercial entities apparently want to sell their existing products, not offer services, even if paid, for conversion based on a library’s criteria
• University of Michigan is digitizing 3000 to 5000 books a year with their own staff. Converting 7 million volumes at this rate does not scale. Entities do exist who could do this work. Trade-off – Regional, local materials won’t get done. Maps, newspapers, rare books won’t be on the lists of most commercial players and many libraries will be unwilling to send these materials offsite. Management oversight is required; could have regional digitizing centers with overseeing body.
• Good result will be materials owned in common by those who support the process.
• If there is help from outside funding sources, the contract must specify that the outcome of the digitization is owned by all.
• Organizations exist already with administrative structures in place, such as ARL, CRL, CLIR, DLF, and a steering committee can offer this project to one of them to move it along
• CLIR may be available, and could work with ARL
• Government documents have to be included, and could be part of a first pass since there are no copyright issues. ARL has looked at a project to digitize them but there was no confidence in its operational skills to manage massive conversion projects. 2.2 million volumes from GPO are available for digitization.
• May not want a library group to process the materials. OCA as a possibility. Toronto project has been effective.
• Create a central fund of money from contributions and have institutions apply for funds to get the digitization done
• Issues exist within institutions – Michigan considers preservation quality v. outsourcing and has different practices in different departments within a single institution. Enforcing common practices across a national project would be much more challenging.
• Standards have to be agreed upon and then followed for all types of digitization – ocr, keying, page scanning, preservation quality – and to decide which process applies to which documents
• Standards have to be set quickly (they should not be studied and delay movement forward) and should be based on well-established existing programs which already have standards in place – Michigan, U VA, UT
• UT outsourced some digitization to another library. Metadata discussions were dragging on and in the meantime Google had already put up those titles
• Google content coming back to Michigan is so much to manage and make searchable that it will take time to become accessible. Google’s practices aren’t necessarily the standard.
• OCA with RLG and Adobe on a platform might be a way to go. OCA is interested in helping as a partner. Confident in OCA but they only at present plan to cover a small number of volumes
• Question from the earlier discussion (at CCDO meeting) regarding statistics: exactly how big would this project be – 1923 and back, numbers of newspapers and manuscripts. Michigan may have some numbers
• Eliminate (or at least defer) that which is already commercially done. Research library projects already adopted. Push information out to public or keep in ARL community? That decision may affect how project goes forward. Could work with commercial entities to try to gain access to materials already digitized if in accepted format. Access would require central ownership by ARL community.
• Look at gaps to fill and work on those volumes first.
• Talk to holders of local archives and to research societies to digitize their materials to supplement library collections
• Regional service bureaus may be created; will always have local activity
• Discussion of in-kind contributions from institutions that have already digitized large numbers of files and have capacity to do more.
• Google does not handle fragile items; those are being digitized by Michigan
• Next three months:
o Get notes on blog
o Sell concept – different interested forums – back to institutions
o Find institutions that strongly support the concept to spearhead the challenge
o Need to have collective action from ARL (Janus target)
• Smaller institutions can still digitize their collections to fill gaps
• Need an inventory of what had been done
• However, if you start double checking you never get started
• Ithaca, CDL, LC – Collaborative works; encourage them to help
• Need a central place for the project to reside with expertise in management and perhaps in management of a digitization project
• Timeline – Who – What Next
o Promote to directors and selector communities
o RFI to some entitites
o Determine if CLIR and/or ARL could host a meeting.
ß Include CDL, Ithaca (JSTOR, Portico), GPO, LC (National Digital Library Project), CRL, DLF, ARL, JISC (Stuart Dempster)
ß Create a strawman proposal
ß Reality check on concept
ß Conceptualize a management structure
ß Determine the outlines of an RFI
ß Identify some basic standards
ß RFP
o Need commitment to funding from this group (show of hands was positive)
o Pull together standards from existing projects
-Edward Shreeves
Director, Collections and Content Development
University of Iowa Libraries