Following is a guest blog post from Lisa Shiota, a student at Drexel University School of Information and Library Science and a staff member in the Music Division at the Library of Congress. She explains how she utilized Viewshare in a digital library technologies class.
I am currently finishing classes towards a post-graduate certificate in digital libraries through Drexel University’s online program. This past fall, for my Digital Library Technologies class, our final project was to create a digital library prototype. After looking at several open source applications for digital libraries, I chose Viewshare for my project.
What was particularly appealing to me about Viewshare was the different ways (or “views”) that the information could be presented. I figured it would be worth a try to see how easy it was to use. I requested an account from the moderator using their online form, and once I was approved, I created a login.
Plan and Approach
My plan to build the prototype was fairly simple, at least on paper.
- Identify the physical collection to be used for this project
- Read through the Help pages to learn how to use the system (http://viewshare.org/about/help/)
- Upload smaller test files to see how the system works
- Decide what metadata to record
- Scan covers/title pages
- Upload data
- Build one “view”
- Test interfaces and analyze results
The items I chose for the digital library project are opera scores by Giuseppe Verdi, a 19th-century Italian composer best known for his operas. The Music Division of the Library of Congress, where I am currently working, has most of Verdi’s operas in one print format or another. Although it seemed somewhat limiting to focus on one composer, I wanted to note the contrasting aspects of the collection. For example, most of the items have text in the original language, but there are some that have been translated into other languages. Many of the scores are the first printed editions, but there are several reprints that are represented as well. There are many items that are in manuscript; these are mostly by copyists who had viewed a printed score that had not been available in the United States and had painstakingly made a handwritten copy to add to the library’s collection. These copies are often in extremely brittle condition; many of the handwritten copies and the first printed editions have been copied to microfilm so that a legible, more durable copy could be preserved and made available to library patrons.
After playing around with uploading different kinds of files, I opted to upload a spreadsheet with the items’ metadata. Much of the metadata I chose to compile in my spreadsheet for the items in the digital collection are standard for library bibliographic records: composer, title, publication information, extent (number of pages/volumes), format, language, and call number. I added a couple of fields for internal tracking purposes: a link to the library’s OPAC record where available, and the shelving number for the microfilm version. The notes for each record are mostly mine, which are basic points I found noteworthy about the item.
I scanned the covers (or title pages, in absence of a cover) of the opera scores on a flatbed scanner and saved the images as .jpgs on my personal webspace on my school server. I then added the image URLs to the spreadsheet.
Lastly, I chose to include certain metadata– preferred title, librettists, and performance dates– solely for the purpose of being able to explore the available Viewshare presentations. I wanted to use the preferred title (or uniform title) of a work so that I could group items that were the same work together even if they had different titles on their cover or title pages. I wanted to highlight the names of the original librettists for searching purposes. I recorded the dates of the first performances of the opera (from the “Giuseppe Verdi” entry in Oxford Music Online so that I could experiment with the timeline view.
My final version of my digital library prototype includes List, Table, Timeline, and Gallery views, as well as facets for browsing by score format, language, and librettists, and is publicly available at http://viewshare.org/views/lshiota/verdi-scores/.
This project taught me a lot about the many components involved in creating a digital library. Based on the results of this prototype, I concluded that digitizing the library’s entire opera collection of several hundred items and making them available through Viewshare would prove to be too cumbersome to do. Other smaller collections, such as the division’s archival collections containing short correspondence, sketches, or photographs would work better here. Viewshare’s built-in interfaces for maps, timelines, and graphs would be great for users to interact with the digital collection in a way that they might not be able with the physical collection.