Reference Reviews
Péter's Digital Reference Shelf
June 2007
Title: JournalSeek
Publisher: Genamics
URL: http://www.JournalSeek.net
Cost: free
Tested: June 29-30
The Context
While huge book directories and catalogs are available in open access format, and some of them (especially Amazon.com) are spectacularly rich in content compared to the subscription-based digital edition of traditional Books in Print product, the worthy open access journal directories are few and far between, even though academic libraries spend a much larger proportion of their material budget for journals and other serial publication than for books.
Hence serials directories are even more critical for libraries but many cannot afford the pioneering work of Gale’s Directory of Publications and Broadcast Media, the much improved Ulrich’s Periodical Directory, the Standard Periodical Directory of Oxbridge (marketed as MediaFinder in online format, or The Serials Directory of EBSCO. This latter, by the way, does not deserve the definite article in the product name anymore as it seems to be left rotting with no new record added in the first six months of 2007, and with a single record about an Italian journal Urbanismo in the past 18 months – although even that single records update date is absurd, and obviously wrong: July, 27, 2007, almost a month later than I am writing this review.
Just a few years ago we had more choices for open access journal directories. I particularly liked PubList, the free subset of the Ulrich’s database, which was limited to the then-active titles, but still was very useful as a digital ready-reference tool. Unfortunately, it was later acquired by Infotrieve, the document delivery company, which has been showing for several years the same message about doing technical developments, but for more than six years I have seen many non-functioning options, such as browsing by journal title, searching by subject, and titles that have changed title once or twice since Infotrieve acquired this directory. For example, it still believes that there is a journal called On-Line Review, published by Learned Information Limited, and edited by Martha Williams, even though it changed its title to Online & CD-ROM Review, then to Online Information Review, is owned by Emerald Publishing Group, and has been edited for many years by Gary Gorman. Searching by ISSN does not seem to be disabled, it allows you to enter the ISSN but then sends the message that "Publist is currently maintaining the data. Your search criteria may not return any result at this moment. The search will be fully functional on Nov 22, 2003. Sorry for the inconvenience!"– the perfect example to see how grossly neglected this once excellent has become and remained in the hands of Infotrieve.
Oxbridge offers a demo version with useful data, but with a 350-item sample it is extremely limited for practical use. The Serials subset of the free Library of Congress Catalog is huge, but its records are slim with useful but barebone bibliographic descriptive data. All these facts make the usefulness of JournalSeek quite obvious.
The Content
JournalSeek is the product from Genamics, Ltd., a New Zealand company, producing several open-access databases, such as SoftwareSeek and GenomeSeek, beyond JournalSeek. JournalSeek has information about more than 92,000 serials from 2,725 publishers (Genamics refers to 2,800 publishers, then on another page to 3,242 publishers. In spite of this convincingly exact number, my tests could confirm only 2,725 titles). Although the number is less than half of the size of Ulrich’s and The Serials Directory, it is still quite impressive for an open-access directory genre, even if I don’t agree with the claim that this is "The LARGEST fully categorized journal information directory on Earth." This is especially strange when Genamic explicitly mentions elsewhere that 37,340 journals are categorized. To its credit, all journals have at least one subject descriptor, another form of categorization. The categorization is appealing, especially in comparison with that of EBSCO, which claims to have Dewey Decimal and LC Classification codes assigned to all its records. This is very untrue, as there are Dewey classification codes assigned to less than half of the records in The Serials Directory (I could not check the availability of LC classification codes), and with all respect to the classic function of classification codes, these days end-users are just not willing to find somewhere what the LC code is for, say, Computer and Information Science to find journals from that discipline. This is right in front of them in JournalSeek on its first page.
Classification is appropriately wide and deep in JournalSeek, and through the category browser is explorable in a visually pleasing way. You need only click on the broadest subject category to see what are its narrower subjects. As for the hierarchy of categories, it can never please all the people, all the time. I find it odd, for example that the journal Information Sciences is listed under both the category of Information Retrieval – along with 23 other journals- and of Information Science when it would be sufficient under the broader Information Science category which lists, which in turn should not list and not exclusively under Information Science, the French pulmonology journal: Revue des Maladies Respiratoires.
The record content is fine, with title, descriptors and ISSN for all the records. Only 7,578 journals have Electronic ISSN. Half of the records have abbreviated journal titles and it possibly should be and could be enhanced, especially when considering its variety in the various databases, or even within a single database, as it is not universally standardized. Several databases and journal archives have free lists of source coverage created by the aggregators and/or by the publishers.
Close to 25,000 journal records have links to the site of the journals, letting the user jump in for further details, or to see the table of content records, as well as at least item level information to learn the article level counts. Many more journals could be enhanced with direct link to the publishers. There are several journals that do have rich sites at the publisher, still the journals are not linked. This is particularly strange in the Computer and Information Science category, where, for example, the excellent Annual Review of Information Science & Technology has a link to the publisher but the equally influential JASIS&T or even just JASIS (the former title) has no links. The ARIST link is not to the site of John Wiley, Inc. which is the publisher, but to the editor’s university web site, so the lack of JASIS(T) link is not unique. While the publisher has very disappointing coverage for ARIST, it has a much better digital coverage of JASIS(T), and the links would be as simple as I show here and here, and they would take the users by the hand to the current and former title of JASIS(T), and show useful information.
In this regard, JournalSeek should improve, as the sites of the publishers covered by JournalSeek do have together millions of open-access abstracts and bibliographic records for their journals, making the links especially rewarding. The list of the top 25 publishers have more than 10,000 active journals alone, and while the links to their site is great, this linking is not yet implemented consistently at the journal title level.
Elsevier alone has about 7 million records, Springer and Taylor & Francis about 3 million each, John Wiley and Blackwell (which has been acquired by the former but luckily the latter still maintains its own site) has 1.5 million records and growing. About two thirds of those records have substantial abstracts that can be easily searched (except for Wiley which does not have a separate index for abstracts but only a combined full-text and abstract index – indicating how little this publisher’s application development team knows about the importance of field-specific indexes).
The Software
Beyond the category browsing, the software offers two search options: Searching by title words and searching by any words in the JournalSeek records. The search in the title index for the term "digital" finds 66 journals, the search in all the fields returns 216 journal records. This list is sorted by relevance rank and split into two parts, one with links to the journals, and ones without links. In the title search, the proportion is 30 linked and 36 not linked – this is what should improve. Although there is a link to Openly Jake but this once-free service now seems to be only partially functioning – at best.
Although the software could be enhanced to allow a search by the combination of the all fields index with a limit filter to restrict the results by language and by publishers and/or with a limit filter to open-access status, and the JournalSeek records should trace down more links to the publisher site, this is still an excellent resource both for ready reference and for the OPAC enhancement in every library.
— Péter Jacsó