November 28, 2005
SML Room 409
Katie Bauer, Matthew Beacom, Meg Bellinger, Dan Chudnov, Katherine Haskins, Julie Linden, Suzanne Lorimer, Kenny Marone, Fred Martz, Roberta Pillette, Kalee Sprague, David Stern, Joan Swanekamp, Rich Szary
Jeffrey Barnett, David Gewirtz
The IAC has moved from a 60 minute to a 90 minute meeting from now on.
VITAL update from David Gewirtz and Jeffrey Barnett, with time for questions
The group also needs to think about topics for future meetings (suggestions?)
1. Digital Preservation Committee – Bobbie Pillette
Bobbie reported that the DPC felt it was being sidetracked by minutiae, and discussed what it ought to be concentrating on, narrowing it down to:
- guidelines and best practices
- recommendations for administrative and business models
- recommendations for technical infrastructure
The DPC is focusing on the last two, and is forming smaller groups to work on them. This should help the group focus, so then approaching Best Practices will be easier.
Meg asked if there were any timetables yet, and Bobbie replied that there is not yet, but there hopefully will be after the next DPC meeting. Meg asked that an update be given at the next IAC meeting, as well as a general update on the committee itself, being sure to touch on how it intersects with VITAL/Fedora.
2. Metadata – Matthew Beacom
Matthew reported that the Metadata Committee is working on the Yale Elements Set, and that their work is being based on responses to the survey. He described the Yale Elements Set as a Common Elements set based on the Dublin Core. The intention is to develop a common set of elements that will work across platforms to increase access. It will also help analyze usage; there are usage guidelines under development. Work is also being done on mapping to Dublin Core, etc.
Matthew stated that Katherine Haskins spoke at the last Metadata Committee Meeting – they are doing similar work. Meg asked if Katherine shared about timesavings, etc., and Joan answered that they had come to an agreement about what was really needed: which elements to keep, which to throw away, and so forth. She also said there are still ongoing talks with faculty for fine-tuning purposes. Some faculty has said that they believe there are not enough elements, but the Arts Library believes 15 (or fewer) are sufficient. Matthew said that some administrative and technical information is unmappable, and the committee is beginning to address that.
Meg asked about the conclusion of the Metadata 101 workshop series, and Matthew conveyed that there were about 70 people at the final meeting. They are still waiting for the final survey data to come back, but were able to report that the feedback received was exactly what they’d hoped for, and they felt the workshop met its purpose exactly, as a good overview of metadata, etc. Meg asked that the survey report be sent to committee members when it comes in.
In response to the question, has metadata been mapped across databases? Kalee said that OAI has been converting MARC data. Joan suggested that Karen and Audrey should be able to take data and map it. Katie asked if someone could explain that a little, for those of the committee who are unsure exactly what it means. Dan explained that it’s hard to know how your metadata will work across databases unless you have to have a testable system. He felt it could be done with Metalib.
Meg asked that Dan and Kalee explore how we could use Metalib to do test our metadata, and to report back to the committee on how it would work. Joan suggested that this could lead us to better understand what we need in best practices, as well as where we should go from here. Something that needs to be tested, Matthew recommended, was the content of fields that can be widely divergent, e.g. a date field can have several different kinds of dates.
To wrap up, Meg asked that Matthew send the Nelinet evaluation report to the group when he receives it, and for Kalee, Dan, Matthew, Joan, and Katie to meet to explore how Metalib can be used to test metadata.
Jeff Barnett said that OAI consistently reports that the two essential steps are to establish a common elements set, and to have trials to test that set. Schools that have done this have also developed tools to aid them, and Jeff thought that perhaps we could borrow/use these, rather than duplicating efforts. Meg requested that at a future meeting Jeffrey Barnett and Youn Noh should report on the DLF workshop.
3. DPIP – Fred Martz
Fred reported that the working group has met and has a good start on digital needs. Consultation and surveying is ongoing, and a spreadsheet is being compiled that shows what is needed and what is wanted. The spreadsheet will be distributed when completed.
There are road trips planned to Harvard and Cornell to compare the work being done and efforts being made.
There will be a digital workshop in the new year, probably in February or March. It will cover workflow and standards in digital production, as well as outsourcing, which can be used to fill in the gaps. Meg asked if would also touch on Metadata, and Fred replied that that is a large part of it, and that coordination (between groups and committees, etc) is very important. Meg then wondered if there is some overlap with DPIP and Course support mapping services, asked someone to meet with Danuta or Kathrine Aydelott to prevent duplication of effort.
4. POG – Kalee Sprague
Kalee reported that the first meeting of the new committee was on November 17. The members of the new committee are largely the same as the old, with the addition of some consultants: Kathrine Aydelott, Matthew Beacom, Audrey Novak and Jen Weintraub.
The group has three areas of focus:
1. Forum in January
2. Reviewing functional analysis
3. Impact analysis (on public, priority)
Several of the group’s members will attend SEKAI in Texas, and will work on Mellon Grant. Meg asked Kalee to please report back after SEKAI.
5. Collections Collaborative – Rich Szary
The Collections Collaborative is working on getting grants. A call for proposals will go out soon; Rich recommended them to provide additional funding where needed.
Julie then asked, regarding DPIP, if there were a matrix of services available, and Fred answered that there is a spreadsheet, which will be updated and amplified. He said that we are focusing on needs first, then will go on to what’s available (not just at Yale). Meg added that after this is done, we’ll map our needs against existing services. David Stern said that they are creating five case studies of faculty and their use/needs, and also said that faculty have been coming to them to offer information. Meg made an appeal to all committees to please coordinate efforts to survey faculty so we don’t alienate them/wear out our welcome and asked David to coordinate.
VITAL Trial Status
(Presentation by Jeffrey Barnett and David Gewirtz
An over view of the presentation follows. Questions/discussion points are italicized.
David Gewirtz began by saying that, issues and problems notwithstanding, he thinks VITAL is a very viable solution.
Trials began in 9/05, it was found that the code was buggy, limited.
10/05: It is recognized that VTLS and Yale don’t agree on the idea of ‘collection’. VTLS says what Yale wants is coming, and the time is spent developing use cases.
11/05: A production version of VITAL 2.0 arrives (2.1 not due till Spring ’06). However, VITAL 2.1 is reliant on Fedora 2.2; and Fedora 2.2 still has not been released. On the bright side, they have managed to create relatively stable version of VITAL
12/05 (12/21 meeting): Collection building activity continues.
1/06: A report will be prepared
David Stern asked about the possibility of testing normalization at this point, and Jeff said it is, also that the new version will be easier to maintain.
We now have four instances of VITAL installed on 2 servers. What we want is one instance with multiple collections. We also need customization.
Importing/Ingesting data is very easy. All you need to do is create a “Track Folder”, and the drag and drop files into collection. They automatically upload, one of the ways VITAL automates some Fedora functions. VITAL also allows you to drag and drop images, URLs, html, etc directly from the web into a collection.
Rich asked if it brings in imbedded metadata, and David replied yes, it does.
Vital has flexible ingest modalities – otherwise you would have to create objects. This is the kind of added value we were looking for in this product.
Julie asked if the final report will estimate cost of using only Fedora vs. using VITAL and Fedora. David said it hasn’t been done, but it can be.
Early on we mapped out use cases of what we expected VITAL to do, for example the Yale Indian Papers Project (YIPP). Using Classes2 you can easily link to documents and objects. Microsoft Word documents are not recommended; using them adds an extra step. Text documents work best and are recommended.
If you search YIPP on a date, it will search the metadata as well as the full text.
One Fedora ‘object’ is comprised of multiple data streams, e.g. images, annotations, spreadsheet, etc. Metadata is used to associate a file with a multipage object. You can also use Adobe documents and VITAL will extract the data from them. Also, the user does not need a VITAL portal to pull an object out; click on a link and it goes to an html page.
Kalee asked if the Committee felt it should set a date, by which we decide to continue or discontinue working with VITAL. David said the Report will be done in January and suggested we discuss it more then.
Suzanne asked if there was there a list of collections to experiment with, and Jeff said to go to ClassesV2 and use “Yale Use Cases”.
Dan said he loves Fedora, and he hopes that VTLS can deliver on their promises, but he also had some concerns. He wondered if the ingest process was that valuable, and wants see more before committing. Katherine asked what the maximum benefit would be, and also how LUNA/Digital Library would intersect, and where does VITAL fit in to our existing framework. Are we building a content management system? Are we working towards one point of entry? Is that even possible?
Meg said that she feels if this product can get us out of the box ahead of the game, then it may be worth it, but noted that we have not made a final decision about any project yet. She said this is an interesting and important conversation that should be continued in January, after the recommendations. She asked the group and Dan in particular to formalize their questions.
David Stern wanted to know about use cases and how the software will be used, asking if the Users would be presenting to the committee. Meg responded that it’s in the future, but it will happen.
Fred reminded the group that Fedora can handle object types that LUNA and Digital Library can’t, e.g., full text, finding aids, which was part of the original interest in it. He also thought Fedora could ultimately replace other functions of LUNA or DL as well. Jeff replied that it’s not an either/or proposition. VITAL and Fedora run simultaneously, or you can dump one and use the other.
The meeting ended with Meg reminding everyone to follow up (topics for follow up have been bolded in this document).
Agendas for the next meeting (12/19) will be sent via email. Please send any suggestions to Meg.