design, query and evaluate information retrieval systems;
A main component of a librarian’s expertise is information retrieval systems or the tasks to keep information managed, organized and retrievable for patrons. Querying the system is a main function of a reference librarian and evaluation of the prospective system must be done way before purchase. Each one of these tasks is part of the purview of a librarian. Some libraries will be large enough to have multiple people fulfilling these roles but a small village library will likely only have one person to handle all the tasks. Therefore, it is critical that all students in the program have some level of skill in each of these tasks. Many courses in the program address the design, query and evaluation of library retrieval systems as sections, or all three processes. The evidence here comes from several classes I have taken throughout the years in the program. Working on these three skills has been deeply satisfying.
Designing a system of information retrieval is the focus of one of the mandatory introduction classes to library science. In this class, there are copious amounts of group work around creating databases. In my introduction class, my team created a database of cat breeds. The intention of the project was to give students an idea on how to create a database from the ground up and actually have something that works. Reviewing the work now, the project’s only point was practice and the introduction of concepts that will be built upon in later classes. If I had to make corrections and improve the basic idea, several of the field validation lists would be changed as well as the technology I used to build the database. The project was interesting to me because it was one of my first large library science group projects where I created something functional and usable.
Last semester in vocabulary design, I had the opportunity to do more teamwork on other organizational projects. This time, we were using XML with Dublin Core as a basis to create a database of YouTube tutorial videos. I feel this project is more complete because the team did informal interviews and a literature review to make the final product more useful to actual people. This step was missing from my previous project. Without a true needs analysis, there is a high probability that the proposed project will not be used because there is no need for it or it does not serve a purpose. The needs analysis was divided between the four of us into two people doing interviews, one doing a literature review and the other keeping notes. My task was interviewing since I knew so many professional and amateur photographers.
We discovered several interesting problems once we became immersed in the work. One of my personal tasks on the project was to break down the tools in Photoshop so each one could be included in the controlled vocabulary. A problem we discovered here is the evolution of the software because new tools were added in each release of Photoshop. My team members all had different versions so there had to be compromise on this aspect of the vocabulary. Given the chance to repeat the project, I would only target versions of the software or allow the field not to be controlled. The other option would be to make the Photoshop version and the tools interdependent so the validation lists can vary. Another problem would have been scalability if the project was to be used for any length of time. Designing usable and scaleable information systems is a challenge. With the Photoshop project, scalability was a large looming question with YouTube’s upload rate of ten hours of video per minute. While a brick and mortar library does not grow at that rate, it is clear digital libraries do grow quickly.
Both of my projects addressed above deal with the design of an information retrieval system. The cat database was a functional database while the Photoshop project was the structure and planning of a potentially large and usable piece of software. The librarian not only designs but also queries and evaluates these information retrieval systems. In my paper on “Measuring Up the Big Three”, I look at a single question and see how each one of the major subscription databases fared. As an intellectual exercise, this was a good project. It clearly illustrated the strengths and weakness in each one. A reference librarian researching a patron’s query must target at the best possible solution. For example, it is nonsensical to search for dog breeds in an ERIC database. Even if the librarian is a rock star at creating complicated Boolean searches, she will not be successful if the search is targeted at the incorrect source. The subscription databases are not cheap so it makes sense to be mindful of the final expense when using them.
Searching and querying is an enjoyable task for every library student I have met in this program. Recently an article on Slate.com discussed this love of searching, querying, seeking and desire. The concept in the article is humans are wired to seek and want instead of being satisfied. “Creatures that lack motivation, that find it easy to slip into oblivious rapture, are likely to lead short (if happy) lives.” (Yoffe, 2009, ¶ 11) The article went on to compare the constant searching to the frenzy a pet cat can get into while chasing a laser pointer. As librarians are constantly querying and seeking out information for their patrons as users, there will need to be a cut-off point. Either that cut-off is from the cost of continuing the search in a subscription database or the patron is satisfied. With my paper on the single question posed to Factiva, LexisNexis and Dialog, it was difficult to know when to stop reforming, rethinking and rephrasing my string. That project allowed me to use the skills I learned in my Online Search class, such as pearl growing, berry picking and following citations.
A librarian would do analysis of a retrieval system before purchasing a solution for her library. While I do not have an artifact directly dealing with analysis of an electronic information retrieval system, the papers submitted here all have aspects of analysis. Every design process should include analysis of prospective software, and there are examples of this in the paper on structuring YouTube tutorials for Photoshop users. The study of the three big databases is an analysis of which one was better able to answer the question posed. In my systems analysis class, I reviewed the use of open source software in libraries. In each of these papers, I used my skills at taking a system apart and looking at it critically. This is the same skill I would use it review a possible information retrieval systems solution. A main point to consider when reviewing a system that is not covered in the above papers is the ratio of precision to recall. It is mathematical equation that proves the higher number of recall means there is lower precision and the opposite is also true.
The skills needed to accomplish designing, querying and analyzing information retrieval systems are enjoyable to acquire and employ. Not all the skills will be employed all the time since librarians have specialties and focuses. However, all the skills are needed to be well rounded in this profession, particularly if one plans to work in a small village library where one librarian does all the tasks.
Yoffe, E., (2009, June 19). Seeking: How the brain hard-wires us to love Google, Twitter, and texting. And why that's dangerous [Electronic version]. Slate. Retrieved September 12, 2009, from http://www.slate.com/id/2224932/pagenum/all
Photoshop via YouTube: Metadata For Video Tutorials
Measuring up the Big Three: How is the law addressing online stalking?
Project Cat Database Team Effort