Serials Solutions Central Search cont.
Back in December I blogged about Serials Solutions offering Central Search (my blog entry). Recently Peter McCracken, Co-Founder of Serials Solutions posted a comment on my blog. He mades some interesting points and I thought it fair to repost his comment in a new blog entry.
--
Peter McCracken said...
Greetings - I just saw your post about Central Search. I realize it's a bit late, but I thought I'd like to respond, if I may. I'm one of the co-founders of Serials Solutions, so I admit to some bias here, but I think there are still some valid points that might be worth exploring.
Central Search is definitely not for everyone, but I'd say I think it's for most people. I agree that it'll be great in public libraries, but also very useful in academic libraries, especially undergraduate libraries. For the advanced, experienced searcher, it'll be a good tool for quickly and easily checking beyond the usual range of resources. I do agree that for someone who is focusing on using, say, Medline, or another complex database, they'll likely want to start and end there.
I can imagine a few cases where I wouldn't want to use Central Search, primarily when the terms I want to use are not supported by Central Search. For instance, if I want to search by the time period covered in an article, as I can in America: History & Life or Historical Abstracts, then I'd want to go directly to the database and search there. Similarly, if I want to search using unique fields, thesauri, or other specialized tools that appear in just a single or a few databases, then I'd want to use the database's structure. No federated search tool is going to be able to take complete advantage of those resources.
That said, if I want to do a quick check of other fields, without searching each database, Central Search could be incredibly powerful. Say, for example, that I'm doing research on a topic that is covered by medical and social work databases. I know how to use Medline effectively, and I do my searches there. Just to be sure I didn't miss anything that might not have been indexed by Medline, I use Central Search to search my collection of social work databases -- or, I just search the entire universe of databases that I can access. Sure, the quality of the results will be diminished, and I'll get all sorts of irrelevant stuff. But who's to decide what's relevant and what's not, other than the person evaluating the results? Perhaps I'll find useful information, even from a general subject database, which I wouldn't have found in Medline. There are dozens of reasons why this might be the case. If I find just a handful of relevant or semi-relevant articles, then the search was worth its time -- there's no way I could have searched all 25, 50, or 100, of my other databases as quickly as I could through Central Search.
Or, perhaps I'm starting to tackle a subject about which I know very little. A quick search of general subject databases will give me an overview of issues related to the topic, so I could have a better idea about what terms to use when I'm using Medline's controlled thesaurus. I can also find out which databases are returning the most and best results, so I have a better idea about where to start my focused and more extensive searching.
Or maybe I'm doing research in a medical library for a lay person, and I know that even if I find relevant articles in Medline, they won't be able to read and comprehend what's published in the medical journals. They may need less technical articles, so that they can better understand the issues. Perhaps they would do best to start there, and once they have that background they can move to the resources cited in Medline. And, it's always worth pointing out that while many librarians are experts at using MeSH headings and developing intelligent, accurate search queries, many of our patrons are not. Of course, this includes not just undergraduates, grad students, and the general public, but also our highly educated professionals. As the saying goes, "Librarians like to search, patrons like to find." If Central Search gets people using the library's resources, and gives them a peek at all that the library has to offer, then it effectively promotes the library and its resources in comparison with Google Scholar, or even just plain Google.
No tool is perfect for every need. There are certainly situations in which Central Search won't be the best tool for getting the best possible results in the least amount of time. But in many, many other situations, and not just in academic or public libraries, I think librarians and patrons will find incredible results from using Central Search.
I welcome any comments or responses --Peter McCracken, MLS
Co-founder, Serials Solutions
www.serialssolutions.com
peter [at] serialssolutions.com
--
Is there anyone else out there who can offer their opinion, perhaps there are some libraries out there who would like to comment on their experience with Central Search?
1 Comments:
During a Google search (ironically) I came across your original post and the subsequent reply. I'm intrigued by the implications and issues they raise:
From the Krafty Librarian:
"I am not surprised that Serials Solutions has created a product like this. Unfortunately the masses demand this kind of product."
"People are lulled into thinking they have searched everything (and searched it well) thoroughly when they search something like Central Search or Google Scholar."
"We as librarians have failed in alerting our users to the pitfalls of such searching."
--
Preface: I'm not a librarian, and so welcome any encouragement to "just move along" if this is a blog by and only for professionals; however I'm keenly interested in the development search tools as an important adjunct to my relationships with public and specialist librarians. I depend upon research using electronic databases almost daily for my work, and I frequently encounter frustration when beginning a search for which there are several valid full-text licensed databases, many which have overlapping content, and needing to repeat the same search ad nauseum and mentally de-duplicate the results across disparate user interfaces. I'm not a MLS holder, but I'm pretty familiar with diverse library e-resources and free/online resources--and can represent the viewpoint of an informed "data/research consumer" here.
My interest in replying, is to respond to what I perceive as a straw man argument and unstated premise in this blog post. It seems your reaction to aggregated/federated searches comes from a concern that such technologies exist to bypass the librarian. That said, I agree the second quote above is frequently quite true - but to blame high-level search aggregators would be to blame the tool for its misapplication.
To a large extent well-trained, intelligent, and skilled librarians and researchers will not soon be replaced by machine intelligence. To ensure that they are not unwisely replaced, the burden however is upon the librarians and libraries to help their clients adapt to new, readily available research tools. How? As you indicate in the 3rd quote above, by educating them why a result is not necessarily an answer, nor may it be complete. They will help even further by using their training, experience, skill and intelligence to help clients identify appropriate research starting points and formulate queries which with better probabiity for good results at each stage of search depth.
Is it truly bad that clients and non-librarians have more direct access to federated searches? With whom lies the burden of educating the end-user of a system on evaluating the quality of search results?
So the straw man that one might extrapolate from your original post is this: that the technology companies are disappointing librarians when they develop tools which can be misused by clients, bypassing the knowledge and experience of their librarians.
More troubling is the viewpoint implicit in "unfortunately the masses demand this kind of product" which may be simply evidence that I've stumbled upon a blog not addressed to me, and therefor I perhaps bristle a little bit at being an unfortunate mass :-) Are the demands of the masses wrong in this case, if they indeed are the origin of the demand for federated and other more "democratic" search tools? I don't think they're wrong, and while I would not advocate that libraries as a policy follow the demands of the proletariat without critical review, I think that our libraries and librarians have much to gain by understanding why we masses are making the demands we are.
Frequently a user, particularly in vertical markets or areas of technical/academic specialization, will be better suited than a librarian to evaluate the relevance of the search results, if not necessarily their provenance or completeness. While a little knowledge is truly dangerous, and we are truly intrigued by a glimpse behind the wizard's curtain, and while we may, with our half-formed understanding, create a major headache for a few librarians--I can't help but think that the noble end goal is better reached by having better educated consumers of more accessible/permeable research tools and their results.
If librarians work with such tools and their non-specialist users (indeed at times with skepticism but not a prior contempt), letting computers do what they do well (i.e., rapidly perform repetitive tasks and manipulate vast amounts of data) and focus on facilitating the clients' use of the tools, technologies such as federated searches will be a boon to libraries, librarians and users/clients.
Indeed, your post prompts the question: how will the librarian and user adapt to an environment where, although the permeability of databases is increasing, the data itself is increasing at a furious pace and the ordering of that data is frequently inconsistent between sources? Should librarians see themselves as custodians and gatekeepers, preventing the spread of corrupt and imperfect results, and if so, how can they "keep up?" Or, shall we continue to view you as skilled councilors and navigators of the ever-expanding sea of electronic resources where the tools to extract valid information are growing at a parallel pace?
Paul McClellan
1000accidents.com
Post a Comment
<< Home