Presented by Marshall Breeding, Director for Innovative Technologies and Research, Vanderbilt University
Marshall Breeding maintains Library Technology Guides site. Today’s topic is next-generation catalogs.
Patrons are steering away from the library. Scarily low percentages of users think to start their research at the library. Libraries live in an ever-more crowded landscape — there are so many places information seekers could go. Our catalogs and sites do not meet the expectations of our patrons. Commercial sites are engaging and intuitive. “Nobody had to take a bibliographic instructions class to use a book on Amazon.com.”
A demand for compelling library information interfaces. Need a “less underwhelming experience” at a minimum.
Scope
Current public interfaces have a wealth of defects: poor search, poor presentation, confusing interfaces, etc. Users need to go here, or there, or elsewhere, to find the kind of information they’re looking for. We make them make choices. The entire audience agreed (by show of hands) that the current state of OPACs is dismal.
We need to decouple front end from the back end. Back end systems are purpose-built and useful (to us). Front end systems should be useful for users.
Features Breeding expects to see in next generation.
Redefinition of “library catalog” — needs a new name. Library interface? Isn’t just an item inventory. Must deliver information better. Needs more powerful search. Needs, importantly, a more elegant presentation. Keep up with the dot com world.
It must be more comprehensive — all books, articles, DVDs, etc. Print and digital materials must be treated equally in the interface. Users must not be forced to start in a particular place to find the material they want. They want information, not format. More consolidated user interface environment is on the horizon.
Search — not federated, but something more like OAI — searching metadata harvested from databases, not just the first results returned by each database. Coordinated search based on harvested/collected metadata. Reduces problems of scale. Still great problems of cooperation. Also — questions of licensing.
Web 2.0 influences. Whatever the next system is, it needs to have a social and collaborative approach. Tools and technologies that foster collaboration. That means integrating blogs, wikis, tagging, bookmarking, user rating, user reviews, etc. Bring people into the catalog. At the same time, important to create web 2.0 information silos. Don’t put the interactive features off on the side — integrate it. Make it all mutually searchable.
Supporting technologies: Web services, XML APIs, AJAX, Widgets. The usual suspects.
New interface needs to have a unified interface. One front end, one starting point. Link resolver, federated search, catalog, web — all in the same place, same interface. Combines print and electronic. Local and remote. Locally created content, and even — gasp — user contributed content.
Features and Functions
Even if there is a single point of entry, there should be an advanced search that lets advanced users get to specific interfaces. Relevancy-ranked results. Facets are big and growing. Query enhancement (spell check, did you mean, etc.) — to get people to the right resources. Related results, breadcrumbs, single sign-on, etc.
Relevancy ranking — Endeca and Lucene are built for relevancy. Many catalogs have default results lists by date acquired. However it’s done, the “good stuff” should be listed first. Objective matching criteria need to be supplemented by popularity and relatedness factors.
Faceted browsing — users won’t use Boolean logic, need a point-and-click interface to add and remove facets. Users will do an overly broad search; you can’t stop them. Let them, but give tools that allow them to correct their “mistake” easily. Don’t force them to know what you have before they search.
Need spell check, automatic inclusion of authorized and related terms (so search tool includes synonyms without user having to know them). Don’t give them a link from “Did you mean…” to “no results found.” That’s rude. Improve the query and the results without making the user think about it.
Don’t get hung up on LCSH — think about FAST. Describe collections with appropriate metadata standards. Good search tools can index them all, anyway. Use discipline-specific ontologies — even if not invented by librarians! — as they are the language of the users.
More visually enriched displays. Make them look nice. Book jackets, ratings, rankings.
Need a personalized approach. Single sign-on. Users log in once, the system knows who you are, and that’s it. No repeated signing on. Ability to save, tag, comment, and share content — all based on the user’s credentials. Allows them to take library into broader campus environment.
Deep Search. We’re entering a “post-metadata search era”. We’re not just searching the headings of a cataloger, but we’re searching the full text of books and across many books. And we can soon search across video, sound, etc. Need “search inside this book” within the catalog.
Libraries aren’t selling things; we’re interested in an objective presentation of the breadth of resources available. Appropriate relevancy for us might include keyword rankings, library-specific weightings on those keywords, circulation frequency, OCLC holdings. Group results (i.e., FRBR). Focus results on collections, not sales.
What we do must integrate into our “enterprise” — university, government body, city government, etc. We need to put our tools out where the users are since (as we know) we’re losing the battle to make them come to us. Systems must be interoperable — get data out of ILS and into next generation systems. And hooks back into ILS from front end.
This won’t be cheap, in terms of money and effort both. But we can’t afford not to make this transition. We don’t have years to study and work to catch up with where we should have been years ago.
Is there an open source opportunity? Yes, but implemented systems are not taking the open source approach, for the most part.
I had hoped for a product review in this session, but the overview of features and desiderata was very helpful. There was a whirlwind tour at the end, but I would have liked an overview of what’s there.
Next Generation Discovery Tools
There was a fair amount of discussion at the recent Computers in Libraries of “next generation discovery tools” — the technologies that, many of us hope, will supplant the now aging OPAC concept and provide better, more interactive, and more…