More google-y thoughts …

November 13, 2009

Sitting at the reference desk has caused me to think even more about Google Scholar, and why it has become my primary research tool (this has just happened without thinking about it …. I’m trying to figure out why). Of course, the easy answer is that I get results, and I get them fast.  Lately, students having been sitting down at the desk, we’ve done searches on vendor-supplied tools, and then as a last check, we skip over to Google.  We then do the same search and become very happy.  Google is absolutely kicking butt when it comes to relevance.  It’s often kind of spooky how well they are managing to get precisely what I want at the the top of the list.

I suspect that at least some librarians will suggest that I’m just not using all of the tools of the vendor-supplied indexes and they could be right. However, if I can just type terms into a box and get the results that I need, I am very unlikely to bother to learn all of the little syntax-tricks etc that are supplied with Ebsco, CSA etc.  That’s just not going to happen.  There are also features of Google Scholar that these vendors just can not compete with.  First, not only does relevant material appear at the top of the list, but it also supplies “related items” as a link, and those items are actually related … this is a thing of beauty and a display of the fact that Google actually understands how research is conducted (by finding the network of related materials, not by endless discreet string-matching searches).  Second, Google supplies the “cited by” link, once again displaying an understanding of just what it is we’re trying to accomplish …

… this brings up and EXTREMELY important point.  When Google is searching, it is dealing with all of the literature available to it, not just the chunk that it happens to control access to.  Thus, they can provide links to all of the items related to and citing the article that is your new favourite.  Whereas everyone else demands that you open new windows and catalogs to search for cited articles, Google links directly to them with a single click, and allows you to continue on your related article search.

Now, my most common advice to students doing difficult searches is to find one good article with a decent bibliography, then follow it’s citations and follow the citations of the cited articles … it doesn’t take long to get a grip on the important works in the field as determined by the people actually doing research on the topic. Vendor-supplied tools do not really facilitate this process … although they supply lots of “hits”, they do a far worse job of describing the relationship between articles and researchers.  Of course, the type of searching that you’re doing in these things would seem to be pretty much strictly literal string-matching.  While the article may have all of the terms desired in it, the student and I quite often sit there wondering what the heck this article has to do with the topic (other than  the fact that the words requested appear in it).  This is the magic that Google has worked long and hard (and successfully) on.  Their search has always been about the relationships between articles … not only that the words appear, but also that the article is “significant” in this area of research because of its relationship to other articles.   Thus, the first step of finding that one good article is quite often done for you … there it is at the top of the list.  All of the connections are then supplied and you’re ready to go.

What this brings me around to is that I’m pretty much done with literal string-matching.  It seems like a primitive and clunky way to do any searching outside of looking for something that you already know the precise title of.  Recent improvements that it would seem have happened at Google (Google Scholar has not always been such a satisfying experience) have made it even more obvious that the “journal indexes” are a second-rate search system.  Even when they work well, they lack simplicity of searching for related and cited articles that are active links to the rest of the literature, and this is what research has always been about.  Google has demonstrated to me that they understand research and are trying to make a tool that facilitates that process, rather than a tool that demands that researchers adapt themselves to the tool.  Vendors, on the other hand, demonstrate that they lack the willingness to work on this, but have a great interest in dividing the literature into little islands of property with walls between them, and then charging you a lot to access their little island isolated from the rest of the world.

The problem is that I can’t quite just ditch all of the Ebscos and Ovids, and CSAs, because they still have the full-text in their limited domain.  Although more and more people are by-passing their front-end and just jumping from Google to their content (through the “Get it at Guelph” button), these folks are going to continue to charge a premium for this content, and may be perfectly happy to de-emphasize search in favour of just controlling the knowledge … this part needs to be solved as well.

As may be obvious, I’m not one of those people who are resentful of Google’s efforts … in fact, I resent the fact that all of the vendors have supplied us with mediocre tools at high prices, and have been quite content with their mediocrity.  I am happy that Google has finally got around to making this situation obvious and at least attempting to create a research tool that is suitable for research … I look forward to their next steps.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: