Lipstick on a Pig: The Sad State of Library OPACs - Roy Tennant
:: Geoff has blogged this already, and I hope the thread is making its way around the library weblog community. Roy Tennant, writing in Library Journal, has summed up, very articulately, the problem with library OPACS:
Recently I viewed a library catalog redesign before it went public. This was the first major change in many years, and it turned out to be quite an improvement to the look and feel of the system. But despite this, it still sucks. Badly.The ISBN example is but one of many. Searching for standards by their number, like an ANSI-approved standard such as ASME A17.2.2-1997, can be a royal pain.
I don't know how much time was spent on this cosmetic facelift, but until the deeper problems that plague this system are addressed, users will remain poorly served. Librarians appear to be afflicted with a type of myopia. We see only minor, easy-to-make corrections instead of changes that will truly affect the user experience. We ask our vendors to tweak this or that to make our lives easier, while the users are left to founder on an interface that only a librarian could love.
One of my pet peeves about the catalog is that we can't keep it straight between fielded searching that is helpful and how and when it gets in the way. For example, nearly every library catalog that offers the opportunity to search ISBN or ISSN numbers requires the user to choose a specific ISBN or ISSN index.
Searching on a number like 1594290202 across the full text of every record in any given catalog (even WorldCat) will return a very small number of hits. So why do we insist that the user specify a particular field? Presumably to allow us to create specific indexes that speed up searching, right? But how hard would it be to extract any set of numeric digits into a generic number index? Then, when someone enters a search consisting of numbers, the number index is searched. This would put the complexity in the back end—where it belongs—rather than in the user interface.
Meanwhile, specifying a certain field often doesn't work the way the user might expect. Let's take author, for example. When you search for books by an author, why do many catalogs return books about that author's work? You guessed it: the added entry. Sure, there are times when users want to get books about that author and their works, but rather than keeping these two categories of search results separate, we nearly always present them in a jumbled mess. Can this ever be even remotely useful?
I've seen many references to Amazon's search engine, which offers a few options for searching (books, dvds, etc.), and when you search the entire site, returns faceted results, allowing you to narrow your search by format, such as those listed above. Even more impressive: move the cursor over new titles, and an "Inside This Book" dialog box appears, that offers options such as:
- Concordance - 100 most frequently used words in the book; click on a word, it will list all occurrences of that word in the book, and list each of them
- Text Stats - data on readability, complexity, number of (characters, words, sentences), and fun stats
- Citations - number of books cited by the book
- Browse Sample Pages
- Search Inside The Book
Libraries with sufficient resources should experiment with other methods of making their collections searchable. High-profile experiments in bibliographic database search systems may help point the way for vendors not eager to perform major redesign projects. A prime example is the Research Library Group's RedLightGreen, a beacon of hope in a sea of library catalog disasters. OCLC is also pushing the envelope, as a recent blog posting by Lorcan Dempsey, the OCLC director of research, illustrates.
So what can most of us do? We need to focus more energy on important, systemic changes rather than cosmetic ones. If your system is more difficult to search and less effective than Amazon.com (and whose isn't?), then you have work to do. Stop asking for minor tweaks from vendors. After all, you can put lipstick on a pig, but it's still very much a pig.