Lately I have been looking promoting and setting up Google Scholar (GS) as a resource for clients. The process of getting GS to show “library links” to full-text of items in a library’s collection in GS search results is actually surprisingly easy. Just a matter of contacting your lin-resolver and asking them to set it up, and the Google-bots immediately begin to scour your page. A lot of libraries (Most university libraries, some government) already do this and GS requires the institution’s log-in to access the full text or it recognizes requested IP addresses and provides direct access.
The love-hate relationships librarians have with Google has been well documented and discussed and the relationship with GS is no different. Since GS searches the whole web for articles, its metadata is usually messy. Too messy by some Librarians account. There have been studies done on showing how it is possible to “game” the GS system and make fake paper and authors suddenly become widely cited pieces that appear at the top of search results. Other Librarians have claimed that Google’s history of using nefarious purposes to rank results may soon bleed over into GS. Will we soon see ads on GS? It has also been worried that GS’s heavy emphasis on the citations in its algorithm makes it hard for new and important articles to be discovered (because older and more established articles will also have more citations and be ranked higher). There also problems with the simplicity of GS’s advanced search.
All this needs to be taken in the context of what it is. GS is a federated search that searches the whole freaking web. I’ve been working with EBSCOhosts Discovery Tool Lately and is it also extremely messy when it comes to metadata. But who can blame them? When you rely on metadata coming from a lot of small individual sources, there will of course be inconsistences with other sources, honest mistakes, and attempts to “game” the system. As a database grows bigger these will get worse, and GS is the biggest.
Also a researcher should only be using this kind of general federated search as a starting point. More in-depth searches that require longer Booleans, subject headings, and citation connections should probably be done on a database like Scopus or a subject specific database.
I am sympathetic with the concerns about not knowing what Google will ultimately do with GS though. That is a worry. Also I would love a better advanced search function and more filters/limiters for the search results. And more clarity about how much citation count factors into result rankings.
The real worry with GS seems to be that it will become the one-stop-shop for research and that it will result in most people using non-peer reviewed and junk science articles that were able to game their way to top of the search results.
A huge and increasing amount of journal articles are now available online. While publishers still charge for these articles, they are often also freely accessible from institutional repositories, author webpage, and research sharing sites like Researchgate.
Google Scholar finds all these freely accessible versions and they appear in their results. I would bet money at this point that out of ten articles I am more likely to find more full-text articles from GS than from my library’s collection.
Also so many researches these days have multiple library collections they can search. (The library collection from the institution they work at, the collection from the foundations / organizations they are a part of, and the university library collection from where they got their degree). I have a friend doing his PHD who uses GS to search two university library collections simultaneously. He set it up so both Universities’ library links appear, and because he works in a niche field, this allows him to find resources quicker instead of needing to check each library’s site individually. Not only does he get full-text links from both of these huge library collections but he also gets all the open access articles that no database other than GS can find.
So wow right?
This is why GS is exciting. But also terrifying. Yes he has direct access to more articles from GS but are these the right articles? Maybe.
There is a lot more written on this topic. I have a bibliography of some articles on GS that can share if people are interested.
I read somewhere that around 70% of articles might now have a free online copy somewhere on the web. I am not sure how true that is. But Google Scholars ability to find and allow people to access free articles that publishers would otherwise be charging them lots of money for is amazing.