NFAIS on discovery services

This post was written by Abe on February 14, 2012
Posted Under: Federated Search,View from Inside

There’s an NFAIS draft Discovery Service Code of Practice that’s up for review.

The National Federation of Advanced Information Services (NFAIS™) is releasing a draft Discovery Service Code of Practice for review and comment by March 16, 2012. NFAIS believes that discovery services have the potential to provide ease of information discovery, access, and use, benefiting not only its member organizations, but also the global community of information seekers. However, the relative newness of these services has generated questions and concerns among information providers and librarians as to how these services meet expectations with regard to issues related to traditional search and retrieval services; e.g. usage reports, ranking algorithms, content coverage, updates, product identification, etc. Accordingly, the NFAIS Code Development Task Force has developed this draft document to assist those who choose to use this new distribution channel through the provision of guidelines that will help avoid the disruption of the delicate balance of interests involved.

I recently got an email raising concerns about the draft paper and why services such as my company’s Explorit Federated Search were being excluded by NFAIS.  I don’t want to go further into the specific concerns of the email but I do want to comment on some things related to Discovery Services.

First, I want to say that, despite all the marketing, the concept of a Discovery Service is not new. Two or three years ago OCLC coined the term Web Scale Discovery to refer to their WorldCat index. Soon thereafter ProQuest started referring to Summon as a Discovery Service and put a lot of marketing muscle behind that. As a consequence of that the term Discovery Service has now become synonymous with a large centralized index even though federated search application such as my company’s Explorit can certainly be considered to be a Discovery Service, and in fact I have been talking about Explorit being a Discovery tool much longer than OCLC, Summon and EBSCO have adopted the Discovery Service label for just what they do.

Large centralized indexes, mostly of metadata have been around for decades now. Consider Web of Science, Scopus, Infotrieve, Ingenta and many others. So OCLC and ProQuest didn’t invent something new, they just developed an improved version of something that’s been around for a long time, stuck a new label on it, and through substantial marketing efforts have gotten the term – Discovery Service – to have become associated exclusively with large centralized indexes.

On the very positive side, for what we do and the issues I have raised about Discovery Services in the past, this NFAIS effort at developing a Code of Practice for Discovery Services (i.e. WorldCat, Summon, EDS, and Primo) raises a fairly substantial set of issues with these Services that Deep Web Technologies can address. I can see this paper as useful in highlighting issues with these Discovery Services. Three big issues that come to mind (I can think of others) are:

  1. Muddiness about who owns content and inadvertent access to content (including meta data). “Lack of user authentication/verfiication and/or the inability to reliably identify an institution’s holdings” is cited by the NFAIS paper as a concern.
  2. The ranking algorithm. Whose content is shown first in Discovery Service search results? There are issues with conflicts of interest among Discovery Service providers that makes me uneasy.
  3. Coverage. What sources are available to users of Discovery Services? Who gets to decide that? What about access to the very specialized sources that Discovery Services are less likely to have access to?

Take a look at the NFAIS  draft Code of Practice and raise your concerns.

Reader Comments

Hi Abe,
Of course the concept of “discovery service” is not new, and the term is loosely bandied about. But whatever these relatively recent services are called, they are popular with library users, and especially librarians, because they look a bit more like a library catalogue than a publisher’s index. The library’s catalogue records are included in search results, and the library’s “holdings” (database subscriptions, journal content coverage, etc., plus selected open access material) are matched to the central index, so the user can retrieve results just for documents they can readily access. It doesn’t provide comprehensive search, you wouldn’t rely on it for researching your PhD thesis. But it’ll do for the inexpert searcher and is seen as preferable to the alternative, using web search engines.
It seems to me that your company’s product fulfils a different, more specialist need, and that the likes of Infotrieve and Ingenta have simply missed a trick.

Written By Laurence Lockton on February 15th, 2012 @ 9:14 am

Add a Comment

required, use real name
required, will not be published
optional, your blog address