Cost: free service
Tested: August 25-28, 2009
EBSCO once again made another important indexing/abstracting database freely available to librarians, other information professionals and the entire public (that cares). It deserves more up-to-date information than is currently offered by EBSCO because its content was good at the start, and grew by 25% since the launch of the database - thus has valid bragging rights. The software has very good browsing, searching, clustering and filtering options, as well as the splendid linked-full-text filter, but there is a software glitch that may deprive those users from tens of thousands of records. who browse and then search by journal names.
The best-ever public relations move by EBSCO was when it made available free of charge the huge Library and Information Science & Technology (LISTA) database with 1 million bibliographic records. It had a captive audience which deserves to be well-served, not only because librarians decide which databases to subscribe to, but also because they can make better decisions if they not only give, but also get, information free. I raved about LISTA three years ago in this column, and my satisfaction kept growing in parallel to the 22% growth of the database.
EBSCO also made K-12 teachers happy with the release of the Teacher Reference Center, which offers more than 800,000 bibliographic records. However, it was not as critical as LISTA, because the much larger ERIC database (with now more than 1.3. million bibliographic records and more than a third of them with full-text content) already has been available free (i.e. not through commercial third parties), while there was no open access library and information science database until LISTA was launched.
GreenFILE may be even more important that LISTA has been, simply because of the much wider public interest in the practice and literature of environmental issues.
I don't want to give the impression that I'm an emerald guy who wakes up every morning thinking about the recycling action plan of the day, and searches environmental databases day in, day out, but there are many who would like to learn more about how to be greener. They will appreciate this free ready-reference tool. Because of the universal importance of the topic, there are many subscription-based databases.
EBSCO itself offers the even larger subscription-based Environment Complete database, with more than 2.2 million records, including full-text articles for about one third of the journals it covers. GreenFILE is a free subset of that database.
Gale offers a huge variety of resources related to environmental issues for school libraries, such as GREENR (Global Reference on the Environment, Energy, and Natural Resources), the brand online portal of text, audio and video reference materials, and the Environmental Studies and Policy Collection of journal articles from more than 700 journals.
There are several indexing/abstracting databases dedicated to the environmental literature, and CSA (now part of ProQuest) offers comprehensive coverage through its Environmental Sciences and Pollution Management "supermarket" database with a total of 1.7 million records. It is a family of a dozen databases ranging from Aquatic Pollution and Environmental Quality to Water Resources Abstracts with databases covering all sub-disciplines of environmental science and technology, including Ecology Abstracts, Environmental Engineering Abstracts, Pollution Abstracts, Toxicology Abstracts, Sustainability Science Abstracts, and Water Resources Abstracts - to name just a few of the components.
The U.S. Environmental Protection Agency has the free National Environment Publications Internet Site (NEPIS) database with a relatively small, but full text searchable database of 31,000 EPA publications.
The U.S. government has dozens of "boutique" databases specializing in environmental information. Those who are familiar with the excellent, very smartly named metasearch and clustering engine, TOXSEEK, developed by Tamás Doszkocs of the National Library of Medicine, don't even need to jump from one government database to another to run a search because TOXSEEK broadcasts the query to a large number of databases that specialize in environmental health and toxicology. Even if several of these databases time out, there are plenty left to lead the users to bibliographic and full-text records on the health aspects of environmental issues, as the results of my search about silicosis illustrate.
In spite of the growing importance of environmental issues, not all news is good news on the related database frontier. For example, Enviroline, the online version of Environmental Abstracts with about 400,000 records, has not been updated since May 2008 on the Dialog system, a key service for librarians working in special libraries. To its credit, this is clearly indicated when the database is selected. The same is true for the British WasteInfo database of 123,000 records that has not been updated by its original content provider since July 2001. It is likely to be removed soon because of the lack of recent content.
GreenFILE is closest to the EnviroLine database in terms of its size and composition. When it was launched a year ago, it had 384,000 bibliographic and 4,700 full-text records, that's why you see this information pop up on the Web most often from contemporaneous announcements, newswire items, editorial comments and mini-reviews.
My tests in late August 2009 show that it kept growing, reaching nearly half a million bibliographic records. It is more appropriate to refer to these records as abstracting/indexing records, as 344,000 records have abstracts. Even this adjective may not be perfect, because 83,000 records are enhanced by cited references. True, this figure represents a low proportion of the records, but it is a worthy feature, added to records for publications from 1964.
GREENFILE has about 5,600 full-text records, but the number can be much higher depending on two things. One is the number and type of EBSCO databases that your library subscribes to; the other is the attention of a savvy systems librarian who provides a link to GreenFILE through the proxy server of the library rather than just directly to the "public" URL of GreenFILE. I'll come back to this issue in the software section of this review.
The number of journals covered by GreenFILE is claimed to be 600, but this is not a very relevant number because well-covered, a great many other journals, such as Biodegradation, Biodynamics, Biological Chemistry are represented only by a handful or just by a single record(s). There is a serious software glitch that may deprive users who search by journal names from tens of thousands of records, as I discuss in the software session.
What is relevant is, that there are records for 293,300 academic and 160,000 magazine articles, 3500 books, 24,600 book chapters and 4,700 reports - showing a good balance in the genre of documents to please a wide population from academics to undergraduate students to laypersons. There are records with various types of illustrations (black and white and color photos, charts, diagrams, graphs and maps), but their number is very small: less than 1,000. Then again, the software trick that I mentioned above, can significantly enhance the number of records with illustrations - borrowed from other EBSCO databases that the library subscribes to.
The database coverage goes back nearly 100 years, but 90% of the records is for source documents published in the past 25 years. This ratio and trend, of course, is far better than the reverse one experienced in the apparently defunct WasteInfo database, which is getting more and more stale day by day for not having a single record for documents published in the most recent ten year period.
GreenFILE uses the standard software that EBSCO developed and kept improving for its textual indexing, abstracting and full text databases. Its particular strength is the wide options for filtering, such as the original, nearly 7,000 search results on global warming, by limiting it to, say, 943 academic, peer reviewed journals with cited references, or to cover story articles from magazines, then to the most current papers simply by checking filter boxes and sliding the time-line gadget.
Browsing the indexes of a variety of data elements is another forte of EBSCO. Although it is considered by man an old-fashioned habit, I still do that because it provides useful quantitative and qualitative impressions of the database, such as the distribution of records by publication years, or the accuracy and consistency of geographic descriptors, or journal names.
Browsing the index of journals has let me discover in this case a significant software glitch. There are many journal names that appear incorrectly in the index, preceded by the letter "a" as in a biocycle for 1,036 records. EBSCO uses all lower case style which is OK, but the indefinite article is not OK. There is a also a correct entry in the journal name index under biocycle for 6,032 records, i.e. without the indefinite article. If the user is browsing like a good boy, to check if the journal appears also with a subtitle (it does not ), then happily chooses the single entry from the index to launch the search, he is deprived from the 1,036 records assigned to the wrong name a biocycle.
This is not a problem when searching directly for the term biocycle, as it will retrieve all 7,068 records for papers in that journal. Of course, you may say, that's why searching without browsing can be better than browsing-and-searching. There is a logic to it, but browsing remains an important step when one palns to search by journals, author affiliations, and other data elements that are notoriously inconsistent in abbreviations and punctuations, and simply erroneous in many databases.
The least known best feature of the software is the meshing of indexing/records with the full text of the article, if one of the full-text EBSCO databases subscribed by the library has the full text version. (This is also applied to the LISTA database, making more than 60% of the indexing/abstracting records to appear with the full text in the result list. In my case, out of the 1.2 million records in the LISTA database, 675,394 records have the full text - a stunning result.).
All that is needed is to list the GreenFILE database with a link through the proxy server (showing that the link to the free database comes from a library that has subscription to EBSCO databases), and the rest is automatic. The linked full-text records are not searched from GreenFILE and the other free databases of EBSCO, but I am not complaining about. I happily accept this compromise.
Here is a simple example to illustrate the huge difference. The University of Hawaii Library does provide a link both trough proxy and directly to GreenFILE (the latter is not really needed, and help the walk-in users who don't have UH library card). For the query about "beach erosion" both versions show 59 hits. However, when filtering the search to records with full text , the direct link alternative to GreenFILE shows the message "no results were found", while going through the proxy route link, Ebsco shows 20 hits, meshing records on the fly from the other EBSCO databases subscribed to by the university when displaying the records.
The Hawaii State Public Library (HSPLS) also has an entry about the GreenFILE database in its directory, but the link is direct not through the proxy, so the filtering of the query on beach erosion produces no result.
HSPLS may not have produced full text for 20 of the 59 hits as the UH library system did, because the latter has some EBSCO databases that the former does not have, but certainly could have produced at least a dozen full text record. Hopefully, HSPLS and all the libraries that currently link directly to GreenFILE http://www.greeninfoonline.com , will offer the link-through proxy.
It is a far better solution because even with this smaller file, the total number of records with linked full text is 98,500 at UH (as opposed to the 5,627 full text records, that come built in GreenFILE. Once again, the total number of on-the fly full-text "upgrades" depends on the number, type and edition of EBSCO full-text databases (Academic Search Premier/Elite/Complete, etc) subscribed by the library. This is an excellent idea, a perfect example for the "good for the goose, good for the gander" principle.
I have been arguing for long for this practice time and again with systems librarians, and sometimes I was successful. HSPL switched the direct link to LISTA through their proxy, but did not apply the same principle for GreenFILE - yet.
GreenFILE - along with the other free databases would deserve more PR, and widely publicized URL. Oddly, when one uses GreenFILE directly, and clicks on the Change Database button, no other databases are shown, missing the best opportunity to alert a user in the heat of the moment that there are other two free databases LISTA and the Teacher Reference Center.
On the other hand, others, who have nothing to do with the creation of GreenFILE (and the other free databases offered by EBSCO), may send a misleading message. For example, it is very disappointing that SLA makes it look on its Web site as if GreenFILE would be available free only as a special perk for SLA members. If you click on the button to access the database you get an error message informing you that the link is a member-only page (and shows a DowJones logo - sponsoring it). This is also a disservice for the SLA members who don't have their SLA log-in data while away from their office at, say, the Internet booth at a conference, or have not yet memorized it.
EBSCO is the one who is really and smartly managing and sponsoring the database, and you don't need to be a special librarian (in either sense of the term) to access it free - and that's what should be communicated loud and clear.