Ovid has been a trail-blazer in the development of many online search functions: including the best full-text rendition of articles based on SGML more than a decade ago,  the best solution for mapping query terms into controlled vocabulary, and for the visual display of thesauri, and the first Java-based search engine in 1998. Many of its novel solutions showed up in my Picks and Pans column (always in the Picks section, except one time, when its implementation of the Books in Print database landed among the Pans). It also won my cheers earlier.

I put it through it paces using  the Library Literature database of H.W. Wilson, because of the chance to enhance tens of thousands of indexing-only records as shown below - with abstracts from the open access archives of publishers, which publish library and information science & technology journals covered by Library Literature. Without enhancements, the Complete Reference link in the result list takes you to the complete record whose completeness could be disappointing these days when users expect at least an informative abstract.

LinkSolver provides the tools to create scripts which generate links to the open access abstract collection specified in the link, and adds to the short records in the result list a label (in this case Peter’s Abs/FT)  to indicate that abstract and/or full text is available.

Clicking on this link takes you to the Wiley Interscience Archive where the abstract is displayed for anyone. The other options are available only for subscribers.

The process of creating the script is very well thought out,  offering a plethora of algorithmic options, and conditions for generating the links. For example, you can define the date range when abstracts and/or full text is available for the journal. JASIS goes back “only” to 1986 (volume 37) digitally. For the testing, I chose to create an entry for Wiley Full Text (even though only JASIS and JASIS&T [the new title from volume 52, 2001] are covered by Library Literature..

As for the query URL, you define placeholders which will be filled in by actual values extracted from the record by LinkSolver. In this case it was simple, because the Wiley site can be searched by a combination of ISSN, volume and page number. Such a search  produces the most unambiguous known item search with one exception – if two items (like a letter to the editor and the response) are on the same page. This is not a concern. One of the few weak points in LinkSolver is that I must scroll in the inexplicably narrow window where I enter the script to see the query in its entirety.

You may create a script for several journals in one fell swoop, as I did for 11 titles published by Emerald (formerly known as MCB University Press), simply providing the journals, and the year range when I want the link to be generated. This is an important option because, in this case, Emerald does not have the entire run of its journals in digital format but only from the year  when they were acquired by Emerald from the original publishers, such as Aslib or Learned Information Ltd.

Emerald makes it very easy to generate links, and deserved cheers for it. It could have been even easier if H. W. Wilson would have included the Digital Object Identifiers (DOI) and/or the Serials Item Contribution Identifiers (SICI). Luckily, the SICI can be reconstructed from the data element from the indexing record, and LinksOvid worked like a charm in extracting the values from the placeholders in the script to create a SICI. 


In all the above cases the abstracts are free for anyone and from anywhere. The full articles are available only from an authenticated computers, i.e. either from campus or through a proxy server. In the case of some journals, such as the open access Web-journals, (Ariadne,  D-Lib, etc.) the full articles are available without any conditions. 

The script is a but more difficult to write because you have to reverse engineer the query syntax and elements but I was lucky here because for my PolySearch Engines I have been doing exactly that for quite some time. With Web-born journals your query may not be unambiguous enough to yield a simple response because of the understandable lack of such elements known form the print world as page numbers, and sometimes ISSN.


Of my test title D-Lib items could have been searched unambiguously because publications carry a DOI (although not for the entire run). As opposed to the SICI, the DOI cannot be re-constructed from the bibliographic data elements, because it is of free structure following the part which identifies the publisher and the publication, i.e. a DOI is not guessable. D-Lib for example sometimes uses the month year and first author last name format, and sometimes the month year and the type of the item, such as 10.1045/december2003-editorial.

Ovid provides a number of predefined scripts for several digital archives which can serve as a model, although the help file itself is very good, and the software gently guides you through the scripting process. I could have used the script for PubMed Central titles but I was so elated by the availability of the entire run of the Bulletin of Medical Library Association in November as open access full text archive  (the Journal of MLA has started as open access) that I created a script immediately to celebrate  the occasion.

 There are a zillion other smart features, such as checking the validity of an ISSN when you specify journals, and rejecting it if the Modulo 11 based check digit in the ISSN does not match the one calculated on the fly from the first seven numbers you entered. I rarely see any serials system which would offer such an essential function. There are a number of options to choose for author names which is a godsend as some databases use only initials for the first and the middle initial, others spell out the first name, therefore extracting only the last name turned out to be a good choice when I activated the script for a number of databases.

There are a few things I would like to see improved, but the product is already utterly empowering.  Once again Ovid launched  an innovative, powerful, smart, and exceptionally well designed software, which earned its place among the cheered ones.

It is not inexpensive, but it can enrich  the indexing and abstracting databases licensed by the library so well that the investment will soon pay for itself (not necessarily in a tangible manner) by alerting users of the availability of the articles in full text in one of its  digital collections, and providing one click access without the  need of going to  the online catalog and look it up.

Back to Cheers & Jeers 2003