Book Search & Catalog Search

Scott Frey, Reference Librarian, Western State University College of Law
Maria Armitage, Digital Experience Analyst, Digital Services, Columbus Metropolitan Library
Amy Barnes, Digital Experience Analyst, Digital Services Department, Columbus Metropolitan Library

Scott Frey:

• Question: Why would a librarian want to develop a search engine for the library?
  Answers: In order to collect resources from various websites onto one site that are customized for your library users. Add public domain or other sources (like google books) to create library resources for customers. Collect resources into one site, tailored to your individual library, using open-source search engine programs.
• Technology issues: what is difficult in setting up a search engine?
• Answers: Insufficient hardware, insufficient operating system, insufficient bandwidth, buggy software, file formats that are difficult to index (for example, pdf files), restrictive IT policies, computer and network security, administrative skills of staff to create search engines.

Search engine software examples:

Nutch Apache Software Foundations
problem: hasn’t been updated in a year. Most recent: 2 April 2007: Nutch 0.9
Scott Frey set up a search engine using Nutch for a small project at the law library where he works.

Thinking about developing a search engine yourself?
Here are some sources to read:
“A comparison of open-source search engines“
“Why writing your own search engine is hard“

- – - – - – - -

Maria Armitage and Amy Barnes, Columbus Metropolitan Library
This session was about the transition to a new opac using Aquabrowser, a change from the inhouse catalog search that they had used previously. They originally created their own opac in 2000.
link to the new catalog: http://catalog.columbuslibrary.org/

This catalog uses Aquabrowser that draws data from the original homegrown catalog.
Observations from results of change to new catalog opac:

• Their library customers were accustomed to the old catalog and title searches; the new catalog didn’t yield the same results in quite the same way for customers who didn’t know how it worked.
• Need to fix problems fast!
• look at stats to find out where the problems are and to find out how well the new catalog is accepted and used.
• MARC database is more visible for duplicate records, errors, etc.
• when they launched the new catalog, they found out that their customers were library users, not necessarily web users.
• New projects: database integration using federated search tool, social networking (adding tagging, ranking and reviewing), library events integrated into the catalog: add events to author entries, kids catalog, syndetics content info.