I have always been one of those people whose technology experience is just enough to get me into trouble. Coming into graduate school, I was really excited to push myself to go out on a limb and gain some new, in-depth technology experience. Two technology projects that I took on have been especially challenging to me.

When I accepted my position as Grants and Funding Information Services (GFIS) Manager at Suzzallo Library, I knew that one challenge would be re-designing the web site. While I do have experience coding HTML and using Dreamweaver software, I have never approached a web site re-design from a usability perspective. The GFIS web site is currently a wealth of valuable information - that is poorly organized and hard to navigate. I am using feedback informally collected from users during instruction sessions and reference interviews, along with my coursework in knowledge organization, classification and thesaurus construction to create a plan to redesign and re-classify the information on the site. Another problem, apart from the old-fashioned look and poor organization, is the lack of consistency with the rest of the University of Washington Libraries pages. This lack of consistency throughout library departments requires users to re-orient themselves for each library's homepage, creating usability issues.

I also want to utilize Web 2.0 technologies to support the instruction aspect of my position. To begin moving in this direction, I began a GFIS blog in January of 2008 and will be implementing Meebo chat services with the re-design launch. The purposes of the blog are two-fold: to increase awareness of GFIS and our services, and to centralize as much information pertaining to funding opportunities at the University of Washington as possible. I continually have feedback from students that there are just too many resources to consult to find out what assistantship opportunities are available and when information sessions for specific grants are happening on campus. To date, the blog has more than 100 email subscribers and approximately 100 feed subscribers. I am hoping that the launch of Meebo will give students better access to reference services with the ability to ask questions when they have them, rather than setting up an appointment or waiting for an email response. What has challenged me the most in these endeavors are the ideas of thoughtful design and usability on the web, as well as careful implementation of new tools. I sometimes feel that libraries are very quick to invest a significant amount of staff time into research and development of new free web tools before really thoughtfully exploring how these technologies can best serve their communities. The web site is currently in design and will be launched in May 2008. You can view the current site and a preview of the new homepage at the links below.

View the current web site
View the beta web site

Note: Since submitting this portfolio, the GFIS site has gone live earlier than anticipated. You can view the official site here:
GFIS web site

Web usability is just one technology challenge I faced working on my MLIS degree. In the winter of 2008, I took a course in the music department on archives for ethnomusicology. That class introduced me to the University of Washington Ethnomusicology Archives and Spencer Lamm, a fellow MLIS candidate. My experience in that class drove me to volunteer in the archives the following spring. I continued to talk to Spencer about a digital library project he was working on for the 911-Media Center. He was using Greenstone Digital Library Software, an open source software program from New Zealand. My volunteer work focused on culling metadata from a collection of 78 rpm records and storing it in a flat excel file. After examining its scalability and ability to handle the multiple file types associated with these 78 rpm records, archivist Laurel Sercombe and I decided Greenstone was a great way to get the 78 rpm collection online.

I was thrilled to be involved with a project that would provide access to a rich collection of music materials that had been hidden in the music building and terrified that I would not have the programming skills to customize an open source product to meet the needs of the user community. Learning how the open-source community works in terms of support and development was an exciting part of this process. I also had to learn how to compromise my metadata schema to work within the constraints of the software. To accommodate all of the image and sound files associated with the bibliographic information in the digital library, each record is constructed as an HTML file. This means that the relational metadata is represented as links in the HTML and not explicitly by metadata properties. Understanding these compromises is an important part of implementing digital libraries because it will impact interoperability with other systems in the future.

Taking on this challenge has made me more confident in dealing with new technologies. With no previous programming background, I had to learn about how the open-source community worked in terms of support so that I could learn how the software worked. Like most open-source software, there was not a comprehensive help center and the documentation has limitations. I was able to take advantage of user- and developer-created tutorials and wiki information to navigate the software package. One specific example involved processing sound files. In the digital library, each item is represented by an HTML file containing bibliographic information as well as links to sound and image files. When I initially built the library, I was unable to figure out why the links to the sound files were not being processed correctly. I discovered that by hosting the sound files outside of the library rather than building them into the collection, this could be solved.

Initially, this was a fine solution, but I soon learned that the department would be moving their servers this summer - meaning that every HTML record would need to be edited to reflect the new server location of its associated sound files. I knew right away that this would be untenable if such a move was necessary once the library was complete with more than 1,000 to update - I had to solve the initial problem and make those links resolve within Greenstone. A member of the Greenstone listserv was able to identify an issue with the configuration of the HTML Plug-in that was processing those files. One quick fix, and the problem was solved. I continue to participate in the listserv to get new ideas from other users and stay informed about the software.

Because the metadata associated with this collection of 78 rpm records has not been previously recorded in any digital format, these items are being cataloged for the first time and it will take a significant amount of time to complete the digital collection. To date, the library only has 10 completed descriptive records. While my official involvement with the project has come to an end, I continue to volunteer and hope to help the archives reach the goal of 100 records in the library by June 2008. Unfortunately, the library is unavailable currently becuase the music library is undergoing a server migration.

See a preview of the digital library
View the documentation for the digital library