Slashdot Mirror
Google To Digitize Much of Harvard's Library
FJCsar writes "According to an e-mail sent today to Harvard students, Google will collaborate with Harvard's libraries on a pilot project to digitize a substantial number of the 15 million volumes held in the University's extensive library system, which is second only to the Library of Congress in the number of volumes it contains. Google will provide online access to the full text of those works that are in the public domain. In related agreements, Google will launch similar projects with Oxford, Stanford, the University of Michigan, and the New York Public Library. As of 9 am on December 14, a FAQ detailing the Harvard pilot program with Google will be available at hul.harvard.edu."
to never leave my apartment.
will there be ads for particle accelerators, scanning tunneling microscopes and tokamaks in the margins?
Google is diversifying extravagently, pretty soon all of us geeks will be driving google cars that can cross reference the library of congress
"Could you put that in a memo entitled, SHIT I ALREADY KNOW!" - Sarge
Also according to the summary, Einstein.
The safest way to approach lava is to have another person with you and he goes first.
I think this is a great start, There's incredible profit here too, universities spend millions for catalogue systems. If I could use one interface to search for books, chapters, and articles on a subject, I could spend more time actually learning, and less time looking at the same damn "no results" page on GeoWeb. Grrrr.
Sleep is for the weak!
If I download a book, when do I have to upload it again? What is the late fee if I forget?
Seeing as Google cached the entire Internet (the last page of the Internet can be seen here): http://www.google.ca/search?q=cache:dQrQDn0dHW8J:w ww.1112.net/lastpage.html+the+end+of+the+Internet& hl=en&client=firefox-a
Google is now looking to cache everything else in the Universe :)
because its time to dive into the deep web. Projects like this are the key to unlocking the vast stores of important which are currently not readiy accessed online. Personally I'd like to see a Google-run free access Lexis-Nexus project.
Please, give me the the values in standard metrics, like Libraries of Congress!
Just how much storage space will all this data consume? It seems like a massive undertaking.
I am ambivalent about this. Will the books be stored as text to enable searching? If so, given that part of a book's character is its font and typesetting, will ALL the flavor of these books really be captured, in the same way that it would be to read them? Something seems likely to be "lost in translation" here.
Currently hooked on AMP
One reason why this is in the interest of big old universities like Harvard is that it will make it much easier to detect plagiarism in students' essays. If published books were included in Google's index, a plagiarism detection service like Copyscape would also be able to check whether content was lifted from printed material, as well as from the web.
About two months ago, Jeff Dean (an employee of Google) gave a talk at the University of Washington about the inner workings of Google. One thing he mentioned was Google Print and how they scan books: they slice 'em up into individual pages, and then feed them through a scanner. This doesn't seem like an acceptable way to archive a library's collection. So, how are they scanning them in? Why not use this method for Google Print?
Ok, so this is just a bit of devil's avocate, but what happens if you just *happen* to have a writing style similar to someone else who was printed before... what if you read something, and unknowingly wrote something in a similar vein in your essay? I assume you could check it yourself, but then that would just introduce extra cost to even write the essay in the first place... or worse, the plagiarists could just "tweak" their papers ensuring that they're "below the radar" by changing enough style to not be recognizeable...
Make sure everyone's vote counts: Verified Voting
Well, there's the Distributed Proofreaders project for Project Gutenberg... but PG isn't a "we must be the source" attitude from what I've seen. As far as PG is concerned, the more eBooks, the better.
DP probably isn't threatened either - they just shift focus to books that are not in the Harvard collection to avoid duplication of effort.
Back around the '60s or so the University of Michigan cut a similar deal with University Microfilms.
U Microfilms set up and ran a microfilming operation in the library system, microfilming everything that wasn't under copyright (and much that was with permission of the copyright holders, such as several large newspapers and many magazines and other periodicals), along with much of the University's records. Rare books, etc.
(If I have this right) the U got microfilm prints of the documents for free and didn't have to pay for the microfilming of its records. University Microfilms made its money by selling microfilms of the various publications (forwarding royalties, where appropriate, to the copyright holders). The rare books, for instance, could now be studied on microfilm with no further stress on the original, and their content became available at many other colleges and libraries. Good deal all around.
University Microfilms was founded by a regent, who was later slammed for conflict of interest. He dropped out of the Board of Regents but the business deal continued.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
For books in Special Collections, they won't allow copies to be digitalized unless they are (1) paid a fee to scan the book (fair enough) and (2) paid a royalty to post the book to the web.
The royalty amounts to hundreds or thousands of dollars per book (about $100/page or image). This allows the libraries to act as a "profit center" for the universities. This policy applies to all UC campuses (I've tried UCB, UCLA, UCI, UCSD).
This is true even though the book is in the public domain (because they have physical possession and nobdy can make copies until you sign a license agreement). This is true even if you're using the book for non-commercial purposes (such as free posting to the web).
Something is wrong here. People donate to UC libraries (either books or money) for the public good. They don't donate so the library can start a business licensing public-domain books.
Despite that, I have been able to scan many books (by using books in open stacks or purchasing them). These books concern Yosemite history and are at http://www.yosemite.ca.us/history/
For older books, most archivists use a cradle and photograph the pages. It's easier on the book, requires no slicing, and there's no scanner to clog with dust.
The disadvantage is the scanner operators need a little bit more training, but that's not a big problem.
December 13, 2004
Dear Colleague,
I am writing today with news of an exciting new project within the Harvard libraries. As all of us know, Harvard's is the world's preeminent university library. Its holdings of over 15 million volumes are the result of nearly four centuries of thoughtful and comprehensive collecting. While those holdings are of primary importance to Harvard students and faculty, we have, for several years, been considering ways to make the collections more useful and accessible to scholars around the world. Now we are about to begin a project that can further that global goal-and, at the same time, can greatly enhance access to Harvard's vast library resources for our students and faculty.
We have agreed to a pilot project that will result in the digitization of a substantial number of volumes from the Harvard libraries. The pilot will give the University a great deal of important data on a possible future large-scale digitization program for most of the books in the Harvard collections. The pilot is a small but extremely significant first step that can ultimately provide both the Harvard community and the larger public with a revolutionary new information location tool to find materials available in libraries.
The pilot project will be done in collaboration with Google. The project will link Harvard's library collections with Google's resources and its cutting-edge technology. The pilot project, which will be announced officially tomorrow, is the result of more than a year of careful consultation at many levels of the University. We could not have achieved a meaningful pilot project without the efforts of the Harvard Corporation; the President, Provost, Chief Information Officer, and Office of General Counsel; the University Library Council; and senior managers within the College Library and the University Library.
A full description of the pilot program follows here, with further materials available on the Harvard home page tomorrow.
With best regards,
Sidney Verba
Carl H. Pforzheimer University Professor and
Director of the University Library
Project Description:
Harvard's Pilot Project with Google
Harvard University is embarking on a collaboration with Google that could harness Google's search technology to provide to both the Harvard community and the larger public a revolutionary new information location tool to find materials available in libraries. In the coming months, Google will collaborate with Harvard's libraries on a pilot project to digitize a substantial number of the 15 million volumes held in the University's extensive library system. Google will provide online access to the full text of those works that are in the public domain. In related agreements, Google will launch similar projects with Oxford, Stanford, the University of Michigan, and the New York Public Library. As of 9 am on December 14, an FAQ detailing the Harvard pilot program with Google will be available at http://hul.harvard.edu.
The Harvard pilot will provide the information and experience on which the University can base a decision to launch a large-scale digitization program. Any such decision will reflect the fact that Harvard's library holdings are among the University's core assets, that the magnitude of those holdings is unique among university libraries anywhere in the world, and that the stewardship of these holdings is of paramount importance. If the pilot is deemed successful, Harvard will explore a long-term program with Google through which the vast majority of the University's library books would be digitized and included in Google's searchable database. Google will bear the direct costs of digitization in the pilot project.
By combining the skills and library collections of Harvard University with the innovative search skills and capacity of Google, a long-term program has the potential to create an important public good. According to Harvard President Lawrence H. Summers, "Harvard has the greate
I happen to work for one.
It's focused on putting otherwise one-of-a-kind materials online for preservation and ease of access, rather than Byron: The Critical Anthology or Cather on the Rye. It's kind of a mammoth, innefficient beaurocracy, though; I don't agree with some of the practices (such as sending texts off to India to be scrivened, rather than just using OCR software), they're very, very slow to incorporate data, and there are a lot of other problems which stem from the fact that most of them are not computer people, but MIMS holders (librarians).
The fact that Google is doing it gives me hope. Hell, maybe I can jump ship.
The library of the University of Oxford, i.e. the Bodleian Library, was the first "copyright" library in the UK - one of only three - which means that it automatically gets a copy of every book published in the UK.
Aegilops
DP probably isn't threatened either - they just shift focus to books that are not in the Harvard collection to avoid duplication of effort.
Are they really going to provide proofread texts? A novel might only take a couple hours to process, but math is going to take hand markup, and some of the more complex critical editions are a bear. Even at only 2 hours a book (and that's not including scanning time), 4 million volumes adds up to 8 million man-hours or a million man-days. At seven bucks an hour that's 56 million dollars. I expect we'll get scans and OCR, but no hand work; there will still be a place for DP. In fact, we'll be better off, with a huge source of scans to work from.
The uncorrected OCR is very useful for indexing (by Google or others), as the 5% or fewer typos are not enough to interfere with indexing keywords. Uncorrected OCR can also be corrected later.
The page images are tied with the uncorrected OCR so you can see exactly what's there.
For an example, see books at University of Michigan's Making of America (MoA) Exhibit, which has thousands of 19th century books and periodicals available.
Only public-domain books will be scanned. In all or most cases the author's are dead. However, this will revive a great body of work and widen access to many.
One class of author may be pissed will be authors who take older works and just slap a foreword or introduction to the front and collect royalties. I've seen this done for many histories. But author's of todays works can count on royalties for themselves, their children, and their grandchildren (if the book is still selling). The copyright term is too long in the U.S., but that's another story . . .
Maybe they could do something really evil to Microsoft, and then we could say: ``Well, you digitized Harvard's library, so we'll let it pass this time.''
See what I've been reading.
It was just a matter of time before a project of this scope got off the ground. I would like to see them team up with Project Gutenberg (and perhaps archive.org) to provide images of the material. Throw in the little transcoder and perhaps wikipedia and we will soon have a killer information resource that can be cross-referenced to silly proportions. This is a boon for research. Projects like this and the public library of science will add much to collective knowledge. It would also be nice to see them team up with the newspaper project! Next stop--public domain LOC!!!
harmonious design
The professor can just wait until the match comes up, and then double-check at that point.
You'd want to do a thorough overview of any potential instance of cheating anyway. A quick run-through would determine whether or not a paper happened to contain an identical sentence clause or three identical paragraphs.
I think the bigger problem would be the second one you described -- that students could plagiarize and then go through each paragraph, changing the wording slightly so as to avoid positive matches. Still, you could argue that this is pretty much what academics is anyway, just with footnotes and a bibliography.
--------
Bleah! Heh heh heh... BLEAH BLEAH!!! Ha ha ha ha...
Screenshot of the service from John Battelle's Searchblog.
python -c "x='python -c %sx=%s; print x%%(chr(34),repr(x),chr(34))%s'; print x%(chr(34),repr(x),chr(34))"
I've been emailing them asking them to do this for years. I'm glad someone is finally doing it! There is only one problem: how do they get past copyright violations? I tried to get Cornell to do this on campus, but they said a lot of their volumes (periodicals, in particular) were still under copyright and hence cannot be scanned. No, it doesn't make any sense to leave these carbon books literally fall apart when we can preserve them forever digitally, but that's the name of the game.
Someone hurry up with nanostorage so I can store the entire content of human knowledge on a postage stamp (with nanosecond seek time and gigabyte transfer speeds, of course)
Call me mundane, but I want Google to index mailing lists, with a nice interface like their "Groups".
If aspiration is a virtue, achievement cannot be a vice.
i'm trying to give up sigs.
It looks like the largest portion of this will be 7 million items from the University of Michigan (compared to only 40,000 from Harvard). Good article from the Detroit Free Press.
No; the reason there are so few copies is there are so few people who want to read specialized journals. And the small audience only accounts for a small part of what many academic journals charge.
No; the problem is not overhead costs or small audiences. The problem is that the owners of much of that kind of content are greedy bastards. There is no reason for the outrageous price of some journals. Some scientific journal subscriptions are in the tens of thousands; even many liberal arts journals are far from cheap. And if you want to copy an article for your students to buy at kinkos, expect them to pay 35 cents a page or more for the copyrights alone.
And many of them are worse than the RIAA in terms of access to content electronically. Journal articles are included in databases sold to some universities You can read articles in some databases but only by loading a .gif of every page one at a time. No copy and paste, no text access at all. So much technology going into preventing the thing from being copied that the online version is actually less useful than the dead tree version rotting on the shelf.
I think this is a great move by Google and Harvard, and I like the idea behind google scholar, but I expect this kind of work to be resisted by many of journals and professional organizations, to the extent that they have in a say in it. This will be a huge boon in terms of the availability of public domain resources, but unfortunately outdated perspectives on intellectual property are likely to hold back real progress for something really useful to scholars in a systematic way. At least until those perspectives change significantly.
For what it is worth, there was an article in the Painted Lady about it today.
In a related story, the IRS has recently ruled that the cost of Windows upgrades can NOT be deducted as a gambling loss.
Actually we do save the images. Many of the initial projects images are saved on CD's but anything from the last few years will make its way to the 'Open Library System' which is an image archive of the DP page scans. You can find a pre-alpha version at: http://www.pgdp.org/ols There are images for about a 1,000 projects there with many more pending me getting around to importing them. Lots of work to be done, developers welcome. Charles Franks Founder, Distributed Proofreaders
I see they've recently added the complete run of the Journal of the U.S. Association of Charcoal Iron Workers. If I'd known that, I could've saved a bundle on gift subscriptions...
FYI, FOIA isn't free, though the fees are pretty nominal. $0.10/page, $18/hr, after the first 100 pages, with a significant educational discount.
The thought of having a spook do my photocopying for me just sounds.... Hrm. Ironic?
What part of "gestalt" don't you understand?