The
meeting mostly focused on mass digitization and the potential Microsoft/Kirtas
project.
Fred
described the LMT/DPIP meeting at which the head of Kirtas, Lotfi Belkhir, came
and explained his company’s process and the Microsoft project.
Re: Kirtas and their scanner and process: Many of us have seen the Kirtas scanner in
action as they visited in 2005. Kirtas
uses a robotic process instead of humans.
There is a lower error rate and quite a bit of subsequent processing.
LMT
was very positive and receptive to his talk.
Points they made:
-digital
quality should be higher for use in classrooms.
-positive
about Kirtas’s interest in quality, not just quantity.
Note
from Alan Solomon during conversation with Fred: this could be an inventory project combined with digitization,
the digital copy adding security. Fred
noted it has been years since we’ve done an inventory in Sterling.
Microsoft
project: they are looking for pre-1923
books in English in certain broad areas (government, American history, English literature,
entertainment…etc.) We noted that
cherry picking will be expensive and the talk turned to selection
discussion. What should we propose to
digitize. Can we do an on-site process
or must the books be trucked to Victor, NY (answer: they can bring their scanners here, we think). What about brittle books? Should we do those? Apparently the project, according to Lotfi
from Kirtas is a “major departure from MS traditional behavior” in that they
are very interested in library collaboration and in developing networked
resources.
The
conclusion at the end of the meeting was that Microsoft should meet with LMT to
discuss the project further [a meeting is indeed planned for late February].
We
then discussed the scale of the project.
Cornell is doing about 120,000 books in a year. Jen noted that we get 100,000 books in a
large electronic database like Making of Modern Law, etc. So if we think in these terms, this project
might not need to be the very large project that overtakes the library.
Can
MS pay for this? For pulling the books,
preparing for scanning, etc? Based on
experience with Ovid, they would probably negotiate to pay for some of our work
but not all.
How
about LSF? Should we scan everything
there? Positive: all barcoded and with Orbis records. Also, would be a help to Access Services
when completed. Negative: low use, not shelved in call number order.
What
about storage infrastructure for huge amount of files? Can the University help with this? Can the Provost fund? Will VITAL/Fedora be ready?
Next steps: AULs and Alice talk to MS in February.
Recorder: Jen Weintraub