Title: RedLightGreen
Publisher: Research Libraries Group
URL: http://redlightgreen.com
Cost: Free
Tested: April 20-21, 2005

The Context

Online public access catalogs are a dime-a-dozen. Very good ones are few-and-far between. Excellent union catalogs are very rare. RedLightGreen (RLG) is definitely one of the rare ones. The underlying deficiency of union catalogs is that although they try to unify the holdings of many libraries, member libraries follow different interpretations of cataloging rules. This defies perfect unification, resulting in scattered records, even for the same edition of the same book.

OCLC WorldCat is not perfect, but it came the closest to consolidating the holdings of many of its nearly 53,500 participating libraries in WorldCat — a Sisyphusian task. WorldCat has some powerful software features (especially the well-browsable indexes and the sort options), but it is not available to the public free of charge. The free subset in the pilot version was rather small, but last October information about the entire collection was made available to Google and Yahoo for crawling. Neither invested much energy in bringing out the best of the WorldCat subset and making the entries prominent in the result list. Although the width of coverage increased significantly, the depth of information in the free version remained the same.

RLG, which is the only competitor of OCLC in terms of online union catalogs, had a different approach. RLG decided to create a subset of the huge union catalog of its college and university members' holding with superb software features. RLG can't make wonders from the heterogeneous catalog records of its member libraries, but it goes the furthest in offering a huge free catalog with an interface calibrated for collegiate end users who don't speak the library lingo and don't care to learn it.

The Content

RedLightGreen is an end user-oriented version of the RLG Union Catalog of about 130 million records for 47 million unique books and other print and non-print materials held by member libraries of the Research Libraries Group. As mentioned, unification is like herding cats. Even with simple titles and author names, scatter is unavoidable. Take, for example, The Stress of Life by Hans Selye. While the layout is simple and not intimidating, it is not clear why there are three entries for the book, items no. 1, no. 4 and no. 5.

At closer look, item no. 4 reveals a run-on subject heading for Stress (Physiology) that may prevent unification. Item no. 5 has the right subject heading, but it (correctly) shows the definite article and for that reason (I presume) it wasn't herded together with the group of 8th editions under item no. 1, even though it is a record of the 1956 edition which is part of the first group. For the seasoned librarian, this is unlikely to be a problem. But the casual user probably does not find it as intuitive as looking up a book on Amazon.com.

However, the additional information displayed, along with the result list, compensates users very well for the possible stress. The result can be narrowed down by choosing one of the options — subject headings, authors or language. With that said, users may still wonder what the difference is between Stress Psychology and Stress Psychological, not recognizing the lack of punctuation signs. The former is really a parenthetical qualifier, i.e. Stress (Psychology), to separate the physiological and psychological stress from, say, metallurgical stress. The latter is an inverse format for Psychological Stress minus the comma (i.e. Stress, Psychological). Generally speaking, the run-on words, lack of adequate punctuation and weird lower-casing of acronyms are too common a distraction.

The terms offered for refining the query are listed by frequency. It would be quite informative to see the number of records (the posting information in professional searcher's parlance) for each subject term and author associated with the topic.

The content of the result list itself is good and jargon-free. The primary subject of each book appears prominently and can sometimes simply reveal the subject of the book if it is not obvious from the title, such as Sleep Deprivation for the book on Physiological Correlates of Stress Induced Decrements in Human Perceptual Performance. It could be useful to link the primary subjects so that the user, looking for a tad less academic treatment of the topic, could get the list of books on sleep deprivation.

RedLightGreen also often offers a group of related links, presumably by passing the books' title to one or more Web-wide search engines and scraping its result list. It could be a bit more filtered, as too often the links include results from paper mills and other less than remarkable sites, but they also include links to first chapters, table of contents and book reviews.

The Software

The software offers a smoother and more rewarding process than many of the fee-based online public access catalog software. Beyond offering intuitive searching by author, title, keywords, subject and ISBN there are two very useful features.

One is the ability to look up the availability of one or more books in a library of your choice. This is not necessarily your very own district or college library, but thousands of libraries that you may choose from a long list.

Why would you not choose your own district or college library? Because it may not offer many of the extra content goodies (table of contents, index page, sample pages and reviews) offered free by another public or college library in another state for anyone. After consulting the much enhanced entries you may easily jump to your nearby library using one of the excellent, free Library Lookup bookmarklets that you may localize.

The other lovely feature (especially for students and faculty) is the option to create a list of books selected from the RedLightGreen catalog and instantly create a bibliography compliant with MLA, Turabian, Chicago or APA. The list can be sorted by a number of criteria, including author, publication year and order added.

There are a few additions I would like to see in the software. One is the ability to sort the search results at least by title, publication year and subject headings. Addition of my favorite feature from the subscription-based WorldCat would also be great: the ability to sort by decreasing order of holdings. This would show which are the most widely held (and presumably best) books on a subject. The result list is sorted by a combination of factors, but it should not be the one-and-only option. The number of RLG member libraries that have the book is one of the five factors for relevance ranking, but it deserves to be a sort (rank) criterion on its own, and it is not the only one I would like to see added.

The number of editions a book had (in a given time span) is prominently displayed and a good indirect indicator for the quality of the book. However, this is not a sort criterion. For example, the book Holy War, Inc. had eight editions between 2001 and 2002, yet shows up as the 20th item on the result list for a search about Islam terrorism, with 2 out of 5 relevance score. Several higher ranked books had a single edition in the past 3-4 years.

Although I am not embracing unconditionally every move that Google or Yahoo make, it would be useful if they were crawling RedLightGreen to its full extent. Luckily, we don't need to hold our collective breath waiting for such a move because librarians are learning about the in-house implementation. The general public could certainly benefit from RedLightGreen materials if they showed up in Yahoo and Google results. Chances of this happening are good, as RLG is very close to booth Google and Yahoo — at least geographically.