COMPUTER SCIENCE CLASS DEVELOPS SOFTWARE FOR CAMPUS LIBRARY
By James Farrugia, Ph.D., Systems and Electronic Resources Librarian
In the fall of 2007, students and faculty in Computer Science, staff from Computing and Network Services, and several librarians joined forces in an unprecedented campus software development project designed to benefit Drew library users. The goal of the collaboration—to improve search results for people using the library catalog—was realized through a software prototype available at http://bob.drew.edu/DrewLibrary. Patrons searching the catalog by keyword can now obtain more useful results ranked by relevance, instead of the default order supplied by the catalog.
The students from CSCI 100 worked with Professor Shannon Bradshaw, Director of the Computer Science Program, and Systems Librarian Jim Farrugia to parse and index the catalog records. “We were looking for a software project with real clients where the students could learn particular programming skills, and also be exposed to the social processes of working with a group of clients. We were fortunate that the Library and CNS were available to work with us on a project that was ideally suited to the content of the course,” explained Bradshaw. “It was stimulating to work with an interested cross-section of the University. The breadth of experience we were able to offer students—including systems programming skills, an acquaintance with the structure of the library’s catalog records, and the creation of a working, useful search-engine prototype—was tremendously valuable.”
Axel Larsson, Enterprise Integration Specialist for CNS, provided the necessary hardware and software configurations so that students could focus on the tasks of parsing and indexing the catalog records. Larsson, an alumnus of the CSCI program himself, commented, “The students were really engaged and enthusiastic about the project. It’s very powerful motivation to be
able to work on something real that is going to be used by and benefit their fellow students.”
Elise Zappas, Humanities and Theological Cataloger, was invited to attend class to contribute her expertise in the semantics of the subject headings, which had to be understood well in order to create an index that made sense and returned appropriate results. Zappas noted, “[The students] asked excellent questions and were surprised, I think, at the intricacy of the
bibliographic record and the vast number of rules required to construct one. I thoroughly enjoyed my visit to the class
and was amazed that they were able to provide us with a successful search engine so quickly.”
Beth Patterson, Reference Librarian, was involved in helping generate the relevance judgments that were used to calibrate the effectiveness of the search engine through its early stages. She was also invited to class to give a librarian’s view of searching. Said Patterson, “Asked to come into the CSCI 100 class and talk about my experiences searching online databases, including the catalog, I was struck by the enthusiasm with which Shannon Bradshaw’s students discussed MARC [MAchine Readable Cataloging] records and their keen interest in how people use our catalog and other databases. The many questions they posed about the information seeking behavior of real life patrons made me realize just how conscientiously they were thinking about the practical task at hand: creating a system that returned catalog results by relevancy.”
The collaboration will be ongoing as the search engine is refined, based on user feedback and additional relevance judgments supplied by the librarians.
(Originally printed in the Spring, 2008 issue of Visions, the Library Newsletter)
