Google scholarly literature search

November 18, 2004 6:23 AM

Stand on the shoulders of giants. Google Scholar enables you to search specifically for scholarly literature, including peer-reviewed papers, theses, books, preprints, abstracts and technical reports from all broad areas of research.
posted by fvw (48 comments total)

Sorry for the pepsibluist google fanboyism, but this is really really really useful.
posted by fvw at 6:24 AM on November 18, 2004

Well it's a better interface and more reliable (in terms availability) than CiteSeer....
posted by PenDevil at 6:28 AM on November 18, 2004

It's got conference papers on, which makes it a lot more convenient than bids.
posted by biffa at 6:53 AM on November 18, 2004

This is such creepy timing. Today, in like two hours, I'm giving a presentation on finding scholarly information sources.

(Twilight Zone music)
posted by Hildago at 7:07 AM on November 18, 2004

This looks really good at first blush. I've done a couple of searches on which I already have well-developed bibliographies, and Google's results are very much on-point.
posted by stonerose at 7:10 AM on November 18, 2004

Makes me want to call a "do over" and go back to college.
posted by shoepal at 7:13 AM on November 18, 2004

Finally, a place to search that doesn't turn up fanfic in its results.
posted by Foosnark at 7:20 AM on November 18, 2004

Foosnark: there's no such place.
posted by fvw at 7:28 AM on November 18, 2004

wow. my old paper from years ago. sniff. the dreams...
what?! only 5 references? the bastards!
posted by andrew cooke at 7:30 AM on November 18, 2004

Is it just me, or is there no way even to append a first initial to an author: search?
posted by Zurishaddai at 7:33 AM on November 18, 2004

andrew cooke, you've just got to cite yourself more - like I do ;) Thanks for the heads-up, fvw!
posted by carter at 7:39 AM on November 18, 2004

Zurishaddai: author:EW.dijkstra works for me.
posted by fvw at 7:41 AM on November 18, 2004

This is great.
posted by troutfishing at 7:51 AM on November 18, 2004

It looks really useful! A few quick searches by author pull up some of the major papers. Subject searching is no better (or worse) than the uncover database, but no keywords.

Even if all they've done is pull together all the databases on-line, that's extrodinarily useful.

CAS# searching doesn't seem to be available, bummer. It seems to think that they're formulae. Searching for compounds by trivial name is really painful. I hope they can fix this. Maybe a copyright thing?

I wonder how the result ordering corresonds to the Citation Index?

Anyway, suuuuper cool! The world has needed something like this since forever.
posted by bonehead at 8:01 AM on November 18, 2004

Name checking this against things I know are out there is revealing some holes.
posted by biffa at 8:19 AM on November 18, 2004

I'll stick with my PubMed and MeSH until they have bookmarking and exporting. The link directly to the journal is nice, but it's just a google search with different meta-data shown, no?

They've been indexing PubMed for a long time; this doesn't seem that revolutionary.
posted by gramcracker at 8:28 AM on November 18, 2004

Yah, it will certainly have a harder time against PudMed/et al, as opposed to the battle for websurfer searches.

Nonetheless, I've been wondering when they'd do this for a while. Looks pretty neat, thanks!
posted by freebird at 8:37 AM on November 18, 2004

fvw, that method doesn't really work that great. The name and initials aren't treated as a single phrase like you would think, but as separate keywords. So if you search for something like author:H.dijkstra, you can see that you get all papers written by a Dijkstra and a person with an H initial. Not quite what you want, unless it's a single author paper.
posted by smackfu at 8:39 AM on November 18, 2004

Is it wrong that I am salivating at using this? Must be one of the things that writing a dissertation does to a person.
posted by naturesgreatestmiracle at 8:45 AM on November 18, 2004

Ah, but there are a few key differences here, freebird. From the Google Scholar FAQ:

This relevance ranking takes into account the full text of each article as well as the article's author, the publication in which the article appeared and how often it has been cited in scholarly literature. Google Scholar also automatically analyzes and extracts citations and presents them as separate results, even if the documents they refer to are not online. This means your search results may include citations of older works and seminal articles that appear only in books or other offline publications.

This is tremendous, as they're ranking pubs on the basis of who references them, taking advantage of the built in networking that comes from having authors reference each other, just as they do with how many people link to a particular page for regular Google.

Additionally, there's the track-back feature (letting you see what articles have referenced a particular article), which some journal publishers have on their websites, but the scope is usually limited to that publisher's family of journals. Google has the potential to link up all journals, regardless of origin. While the index does have holes now, this will undoubtedly change as they (hopefully) cut deals with various publishers and suck down more data.
posted by Mercaptan at 8:46 AM on November 18, 2004

Ahhhh the revolution has begun. Long live Google!
posted by Hildegarde at 9:14 AM on November 18, 2004

"Results 1 - 20 of about 30 for metafilter"

Good ol' thread 25379. I remember it well.
posted by gleuschk at 9:18 AM on November 18, 2004

It's sort of fascinating to see the librarians' take on this (note: a touch of self-linking in there). It's a very curious phenomenon, how libraries react to Google, and certainly their concerns about "appropriate copy" are warranted... But, uh, rather than telling their patrons not to use it, maybe they should making it work in their crappy infrastructures.

Amazing, first day of new memberships and a thread I can actually participate in.
posted by Human Stain at 9:50 AM on November 18, 2004

Although, what is that crap about "Standing on the shoulders of giants". Sounds awfully pretentious.
posted by Human Stain at 9:51 AM on November 18, 2004

Doesn't seem that pretentious, it recognizes that all discoveries that occur now are founded on the research and discoveries that came before. If I can reach up and make an amazing finding, it's not because I'm a giant, but because I'm a schmoe standing on the shoulders of geniuses and giants, and can reach just a little further for it.
posted by Mercaptan at 10:09 AM on November 18, 2004

Very exciting. My "seminal" paper published in Ecography from 4 years ago (before I turned tail and ran as fast as I could from the Sciences) has 9, count 'em 9! citations attributed to it. Never mind that 8 of 9 of the citations come from friends of mine, that last one is proof positive that the WORD ABOUT SPATIO-TEMPORAL PATTERNS IN SMALL SCALE ANIMAL-GENERATED DISTURBANCES IS GETTING OUT! WOOHOO!
posted by mcstayinskool at 10:38 AM on November 18, 2004

Human Stain: It's from a Newton quote: If I have seen further than others, it is because I was standing on the shoulders of giants, referring to the fact that his work built on the work of those before him. Of course it could also be seen as a jab at Newton's rival Hooke, who was very short.

Entertainingly the quote has since been bastardised repeatedly, giving such statements as

• "If I haven't seen as far as others, it was because there were giants standing on my shoulders"
• "If I have seen further than others, it is because I'm surrounded by dwarves"
• "If I haven't seen as far as others, it was because I was standing in the footprints of giants"
• "In computing we step on the toes of those before us instead of climbing on their shoulders"

posted by fvw at 10:53 AM on November 18, 2004

It's very useful, but with library access I still prefer the ISI Web of Science.
posted by euphorb at 10:59 AM on November 18, 2004

Ah, I see. The meeting I just got out of explained the "standing on the shoulders of giants" quote to me as well.

Perhaps I should read a book sometime.

It's very useful, but with library access I still prefer the ISI Web of Science.

That's great and all... but not everybody has web science, and web of science isn't all that useful for "non-sciences".

But, yes, I feel it should be looked at as another tool, not a replacement.
posted by Human Stain at 11:09 AM on November 18, 2004

Just a note, ISI Web of Science also covers:
Social Sciences Citation Index (SSCI)--1956-present
Arts & Humanities Citation Index (A&HCI)--1975-present

Never knew about ISI, thanks fvw.
posted by Mercaptan at 11:19 AM on November 18, 2004

That's great and all... but not everybody has web science, and web of science isn't all that useful for "non-sciences".

The Web of Science (and I think they're trying to rebrand with "Web of Knowledge", now.) includes databases for the social sciences and for arts and humanities. I have no idea how useful they are, though. I'd be curious to hear opinions from people in these fields, as ISI is my personal favorite literature search resource.
posted by mr_roboto at 12:07 PM on November 18, 2004

If you are associated with a university, accessing scholarly databases like Web of Science is extremely useful. If you're not, Google Scholar looks like a good alternative.

I could also see how this could be useful to do some broad-based, preliminary searches over a number of fields. ALL scholarly databases are limited to the journals they have chosen (or are able to) include, while this seems not to distinguish any criteria other than that the papers are on the web. So you might be (effectively) searching a journal about plasma physics and a journal about Haitian folklore studies and a journal about teaching methods at the same time. And every other subject, practically. This is cool.

But it only seems to be eliminating one of the problems with scholarly databases -- that they are limited to data that have already been put into digital form. In fact, this is even more limited, in a sense, because these articles have to not only have been digitized, but put on the web as well.

Also, Google isn't allowing you to do the same sorts of manipulation that some databases do (I'm thinking right now of my current favorite -- the Ovid interface to ERIC). Things like combining search results, exploding or imploding search strings, giving a thesaurus, etc.

This is definitely handy. I don't think it beats everything else put to together, but I'm sure as hell gonna use it as one part of the research for my next paper.
posted by Hildago at 12:34 PM on November 18, 2004

I guess as an undergrad I should be pretty excited about this. Not that research was ever especially exciting. I can use the real library like it's nobody's business, but some of my friends aren't very good. This should be helpful for them. Slightly.

Is it just me, or is Google taking over the world? That isn't even hyperbole. There's a Google agent in my bathroom right now, but he's left his automatic weapon in my kitchen for some reason.
posted by Kleptophoria! at 12:47 PM on November 18, 2004

There's a Google agent climbing up my leg now! And I can see them in the kitchen, eating my wife's jam!

This is actually pretty nifty.
posted by DrJohnEvans at 1:11 PM on November 18, 2004

Actually, Hidalgo, it also includes "non-web" things out of OCLC's OpenWorldcat, which, if your local libraries have contributed their holdings, would show the closest copy of the book to where you are currently sitting.

Now, if they could add a union periodical list, that would also help. I don't know, maybe they have.

If libraries can figure out how to harness this thing to fit into their licensing models, it's really going to be hard on the vendors that create "metasearch" products (such as Ex Libris's Metalib) which do much of the same thing... albeit, usually much more slowly.

It's certainly a scary time if you happen to be both a librarian and a luddite.

Pretty exciting for those of us in library technology, though.
posted by Human Stain at 1:12 PM on November 18, 2004

More from the "Librarians vs. the unwashed masses" front.

In a survey for this lecture, librarians and scientists were asked to name the top scientific and medical search resources that they use or are aware of. The difference is startling. Librarians named Science Direct, ISI Web of Science, and Medline, while scientists named Google, Yahoo, and PubMed (librarians also named PubMed).

I agree with Hidalgo that the ability to search across databases could be nice (it's hard to tell how totally useful this is, yet), but I really like the idea of doing a Google search, then "kicking it up a notch", by clicking on the "Schoogle" link (or kicking it down a notch, the other way).
posted by Human Stain at 1:26 PM on November 18, 2004

Obligatory Ftrain Googlebot link
posted by euphorb at 1:40 PM on November 18, 2004

Is it wrong that I am salivating at using this? Must be one of the things that writing a dissertation does to a person.

Yep. I think I just had something approaching a moment of physical climax using this. *cough*
posted by jokeefe at 1:42 PM on November 18, 2004

At the academic library where I work, we have an add-on feature (SFX from Ex Libris) to our databases that allows click-through access from non-full-text sources to full-text sources. So, looking in Web of Science will not only give you a list of citations, it will give you a tool to potentially find the full-text in another source without re-searching. For now, that's going to make Web of Science, (or the literature database MLA or the psychology database PsycINFO or the education database ERIC) a lot more useful than Google Scholar.

I was also discussing this with a colleague, and we were wondering who the market is? I see a lot of undergraduates who are barely proficient at effectively searching the regular Google. Are they going to even know to find Google Scholar? And are the faculty who are so entrenched in the paid-for databases we have in their fields going to trust going into Google?

Also, as someone mentioned above, for the research that uses some sort of citation management software, the lack of export is another loss of convenience.

What I think this will really be useful for is the gray literature floating around out there, and for Open Source journals.

I can't wait to see what Google Scholar looks like a year, two years from now.
posted by 100watts at 1:57 PM on November 18, 2004

Actually, 100watts, I don't think "trusting" Google is the problem. Faculty are currently finding some of their articles through Google (from commercial sources that the library pays for) and assuming it's free on the web. The reason they are able access these items is because their IP address is registered with the aggregator, but they have no idea how Eresource licensing works, so they assume it's free and they found it through Google. This happens a lot.

As far as the "value-add" that you mention libraries offer that Google Scholar doesn't (i.e. Context-sensitive linking and citation management), this functionality could potentially be added via Mozilla extensions or toolbars or bookmarklets. In fact, I can really see this through a Mozilla/Firefox extension. If the user is unable to get the fulltext immediately, click on a button in the toolbar and a context-sensitive menu appears with options for obtaining this content, whether it be from another aggregator, through interlibrary loan, or whatever.

I guess my point is, libraries need to look at this as something they need to incorporate into their existing infrastructure, not say, "Pah! The user should want to look in 5 completely different database interfaces and our horribly user-unfriendly catalogs to do their research."

As the very quotable Roy Tennant said, "Only librarians want to search, everyone else wants to find."

I guess I have a bit of an axe to grind here.
posted by Human Stain at 2:32 PM on November 18, 2004

I agree that this is pretty good. Although the kinds of scholarly information from an index such as Web of Science (if you have access via a university) is better and includes citation information plus links. But overall GO GOOGLE!
posted by bluesky43 at 2:34 PM on November 18, 2004

This article turned up by gleuschk's search, is a fascinating analysis of the first encounter between David Dark and five fresh fish. Very wierd to see someone getting all academic about the MeFi, but very interesting also:

I find it interesting that both F3 and DD punctuate this episode as a defense against unfair attack. F3 contexts DD’s comments as illogical and sophistic. He views this episode as an attempt to undermine science that he must correct for the benefit of society. F3 also contexts DD’s comments regarding Konolia as support of attacks on vulnerable young women. He contexts himself as a defender of the weak, with a history to support it. DD contexts F3’s comments as at best careless, as worst pointedly abusive. He contexts F3’s arguments as sophistic and stealthily manipulative, and feels compelled by his own self-image as a righter of wrongs to draw F3 away from what he punctuates as an unfair and unprompted attack on Konolia.

I do feel sorry for people who use "context" as a verb.
posted by eustacescrubb at 5:29 PM on November 18, 2004

Here's another librarian take on it. I think it's interesting as a front end fast search with a simple familiar interface, but truth be told it's not really giving full text access to much more stuff [and I don't know about you, but my search was cluttered with books which weren't available at all and cites to a bunch of things that I then had to go to a library and get anyhow, or to a full-text database]. Just like OCLC Worldcat shows you a much smaller subset and says it's showing you "libraries" this is showing you a small subset and calling it "scholarship." Once Google is willing to pony up and explain how they decide what's scholarly. Don't get me wrong, I like it lots, but it's as interesting for what it doesn't do as for what it does. Federated searching with simple interfaces are where things are going and if Google can be the kick in the ass to get libraries going on it [or solve the problem for them in a total, not partial, way] I'm all for it.
posted by jessamyn at 5:43 PM on November 18, 2004

Librarians named Science Direct, ISI Web of Science, and Medline, while scientists named Google, Yahoo, and PubMed (librarians also named PubMed)

Makes sense to me.

A librarian is going to be approaching a question or topic from the point of view of someone who doesn't know beans about it, but wants to find out about it.

A scientist is going to know a lot of the literature already, which creates a different kind of search. A search engine that returns the classics of the literature every time is great for a librarian, and useless or actively counterproductive for a working researcher. Also, things like web of science won't normally have conference papers or working papers, which are likely what you're interested in if you're an active researcher in the area.

I know my first search strategy is to find something recent on the topic, often by googling for a recent conference paper, and look at its references section.
posted by ROU_Xenophobe at 5:48 PM on November 18, 2004

This is a great achievement. Now physicians like myself can really use google's power to know which are the most relevant articles in any subject. Before this it was impossible...

Leo
posted by leof at 6:30 PM on November 18, 2004

Human Stain: You've been able to get OCLC through Google for a while. You can actually search a lot of databases through Google, you just have to enter the name of the database along with the search query.
posted by Hildago at 10:50 PM on November 18, 2004

Hard scientists value search engines more than librarians, but also more than social scientists, or humanities researchers. They also do less research in print materials than anybody else (Engineers do the least amount). It's not because they're bad at researching, or because they're smarter than everybody and know that Google is the best source, it's because if you're doing scientific research on the bleeding edge, nobody has written papers on your topic yet, and your best sources are web sites of other research groups.
posted by Hildago at 10:56 PM on November 18, 2004

Human Stain - your points are well taken. I think Google Scholar has the potential to be something very, very interesting and useful. I just don't think it's quite as good yet as some of the subscription databases out there. I think what you will see, if Google Scholar catches on, is a re-thinking of some of the features we put on top of our databases and how we can make that work with Google Scholar for our end users. I'd like to see libraries and the people that supply products to libraries work with this sort of thing rather than against it.

I know one complaint I frequently have about our databases now is that they don't work like Google. So many of our patrons are familiar with the way Google works that it would be so much easier from an instruction standpoint if we didn't have to make them unlearn what works on Google in order to be successful in the interfaces they're forced to work with on some of our resources.
posted by 100watts at 6:53 AM on November 19, 2004