Week 9.

Last week Jane, Jen, and I each encoded another finding aid and met to merge our mark-up methods.  I think we have the header down now and next we will make sure the <archdesc> (archival description) is consistent, too. Then we can create a solid template that will help John tailor his Perl script and which can be used for the creation of future finding aids.

We also revised our design for the finding aids search interface and mapped out how the search/browse functionality will be tied to EAD tags. Then we met with Sue, Brian, Nicky, and Linda to go over the design and the functionality we are hoping to achieve. Everyone seemed pretty positive about what we can accomplish in the next few weeks.

This week I will be marking up more finding aids and adding controlled access points to them. (We are marking up the finding aids that already have a MARC records with existing access points first.) These access points will be used to aggregate the finding aids into topics and will play a role in how the finding aids can be searched and browsed. I will also try to import them into Archon so that Sue and Brian can use them to work on the php search/browse interface.

Weeks 6-8

During the last few weeks, we have been focusing on coding a few sample finding aids by hand to create a general template that can be applied to the rest of the finding aids - hopefully with a script like John’s. The trickiest part so far is deciding how to handle the box lists. They are all pretty different and very hierarchical with several nested layers. I think we’ve finally come up with a style of tagging that can easily be applied to all of the finding aids, but it may sacrifice some complexity.
We’ve also been busy desiging a new search interface for the finding aids that will allow users to search across and browse all of the collections. I am a little disappointed we won’t be implementing the search interface ourselves, but we need to focus on marking up the rest of the sample set of finding aids with the time we have left this semester. Hopefully we can plow through them and still have time to experiment with XSL and CSS.

Week 5.

The TEI workshop went well. I wish there was more instruction on XSLT rather than CSS, but there wasn’t enough time to really get into it. I’m now in the process of creating a finding aid in oXygen using the EAD DTD.

Good news: John’s EAD script was successful. The box list isn’t quite right, so I’m trying to create a more accurate model in oXygen that John can use to improve his script. Below is a sample of the results of his script.

Week 4.

This week I converted seven finding aids David McCartney suggested as representative of the University Archives into tab-delimited files and dumped them into Excel. I tried to move the fields around to make the content in each column parallel, but it’s difficult because, even when the basic content areas are parallel, the level of detail in each collection is so variable. Meanwhile, we met with John Osborn, Systems Specialist in ITS, to discuss the feasibility of converting the finding aids to EAD directly from their current html format with a script. I was pretty excited to meet with him because using a script seems like the most efficient route for mass conversion. A script could be designed to correlate keywords (like “Scope and Content Notes” or “Biographical History”) with the appropriate EAD tags (<scopecontent> and <bioghist>). And I do get excited about Perl.

This weekend I am going to a workshop at University of Illinois on another XML DTD, Text-Encoding Initiative (TEI). It will be really helpful to get some training on oXygen, an XML editor I’ll be using for the EAD project. When I return, I’ll try creating a finding aid from scratch in EAD. Hopefully that will give me a better understanding of the tags and hierarchy of EAD.

Week 3.

This week we met with David McCartney of University Archives and Janet Weaver of Iowa Women’s Association to learn more about the finding aids and work flows of their departments. I think the most challenging part of this project will be figuring out how to do a mass conversion of finding aids to XML from a wide range of HTML formats. Some of the finding aids we’ve encountered so far have tables, some are multiple linked pages, and even within each department there are several different formats. We have asked each department for a sample set of finding aids to experiment with.

I also spent time this week reading the EAD Tag Library to get a sense of what the goals and possibilities of the EAD Document Type Definition (DTD) are. I also created some sample finding aids in Archon, an EAD editor we are considering for the project. Next week we will assess the sample finding aids from University Archives, Special Collections, and Iowa Women’s Association and try to dump them into EAD.

Week 2.

Last week I did some background reading on the history of EAD and I did some research on the implementation of EAD for finding aids at other institutions. I found an interesting article called Creating the Next Generation of Archival Finding Aids about the Polar Bear Expedition Digital Collections at The University of Michigan. In this collection, users can comment on finding aids and get recommendations based on other researchers’ browsing behavior. The goal of the project was to push the XML further and enhance the online archive experience by trying to emulate some valuable elements of paper-based finding aids (such as dog-eared and annotated pages). This is not a direction we will really be exploring with our finding aids because, as The University of Michigan librarians admit, user-added content makes the most sense for a tightly connected digital collection. Most of the archives our finding aids point to are not digitized and our primary goal is to test the viability of converting all the University finding aids to EAD and to help establish procedures for creating new finding aids. But it’s always interesting to think about future possibilities.

On Thursday Jane and I met with Special Collections librarian Greg Prickman as part of our fact-finding survey. We had prepared some questions about the complexity, format, and workflow of Special Collections’ finding aids. It was helpful to hear an archivist’s perspective on EAD. We will be meeting with librarians from University Archives and Iowa Women’s Archives (IWA) soon to continue our survey.

On Friday, I played around with some EAD code in TextEdit by trying to dump the content of an IWA finding aid into some sample code from The University of California’s Mildred Davenport Dance Program. On Monday I will try out Archon, the program that Digital Library Services (DLS) is considering for the practical implementation of EAD. From Archon’s web site:

“Archon is a web-based tool for archivists and manuscript curators. It automatically publishes archival descriptive information and digital archival objects to a user-friendly website. With Archon, there is no need to encode a finding aid, input a catalog record, or program a sytlesheet. Archon’s powerful scripts will automatically make everything in the system searchable and browseable on your repository’s website!”

They make it sound easy. I look forward to trying it out.

Week 1.

This semester I will be working on a new project in Digital Library Services (DLS). I will work together with Jane to research how Encoded Archival Description (EAD), an XML language, is being used by other institutions to mark up archival finding aids and then convert about two dozen of the University’s finding aids from HTML.

Jane and I met with our mentor, Jen, on Friday. Our first assignment is to find 3-4 articles about EAD and write a summary to share with each other in our next meeting. We will also review the style and functionality of other institutions’ EAD sites and, finally, survey the UI finding aids to find out how many there are and how they are formatted differently in HTML.

Virtual Writing University Project Wrap Up

I did not finish digitizing the Peter Nazareth collection by the end of the semester, but I was able to spend some time on the project over winter break. I digitized about 18 hours of tape. There is still about 30 hours left to digitize and I’m disappointed I wasn’t able to finish it. I guess I was too optimistic; with travel and an illness I just ran out of time. Nick and Diana will continue to work on digitizing, metadata, and the database while I move on to a new project for the spring semester. I hope to be able to return to the VWU project over the summer if there is still work to be done on the database, as I am taking a databases course this semester.

Week 15.

This week I focused on digitizing and writing metadata. It’s hard to record metadata while a tape is being digitized because you can’t pause it or rewind when you miss something. So I’ve been listening to a recording I’ve already digitized on my headphones and taking notes while another recording is digitized in the background. It slows down my metadata progress because I’m basically listening to two recordings at once, but I think it’s the most efficient way to get everything digitized. The “done” shelf of tapes is full now, so we are making progress. I think we’ll be able to digitize everything by the end of the semester and I plan to keep working on the metadata over winter break.

Week 14.

We finished the NonFictioNow project yesterday! You can see the page with the audio links here. We are hosting it on the iwp server rather than modifying the existing page on the English department’s server. To do that we had to download the page we needed from their site without losing any of the images. I hadn’t done it before, but after some help from Jim, it turned out to be really easy. We opened the page we needed in Firefox and saved it onto the iwp server in the “Web Page, complete” format to save the images, too. Doing it this way automatically redirects the image links to the new place where you’ve saved them. Because they had a background image imbedded in their css file, we had to manually save the background image and put it with the other image files and then change the background image link in the css file. This is all pretty basic stuff, but it took us a while to get it right. Adding the actual links to the audio m3us only took a few minutes in Dreamweaver. It would have been better to modify the actual page on the English server because it’s kind of fragile this way and any changes made to the original page won’t be automatically changed on our page. I’m not sure why we did it this way, but maybe after our fine work on this project, the English department will be happy to unleash SLIS students onto their server for the next conference.

Today I got cozy with the Peter Nazareth recordings again. It was nice to be back in the Bowman house. Diana and I set up a Google Doc to help keep track of the status of each recordings: whether or not it has been digitized and converted to mp3, who is in charge of the metadata, and any notes about it. I also ftp’ed some of the finished mp3s onto the iwp server so that we can put them onto our iPods to help us accelerate our metadata progress. These recordings are so rich, it’s hard to balance efficiency and detail when tackling the metadata. I realized the first recording (the 2 1/2 hour Tlali recording) was taking me about an hour of work to capture 20 minutes of audio. At that rate I’m never going to get through these, but hopefully I can improve my pace on the shorter recordings (30 min).

Next week I plan to get at least 3 recordings done and to do more research on existing metadata languages and OAI. I’d also like to talk to Nick about his progress with the database for this collection.