National Library of the Netherlands discloses its Google Books Contract

After the successful disclosure of the agreement between the British Library and Google Books on the basis of the Freedom of Information Act, the National Library of the Netherlands (KB) also disclosed their agreement with Google Ireland today. Albeit the director of the KB tweeted a day ago that not all public information needed to be available on the Web, it was decided to publish the agreement on the Web since there were two WOB (a Dutch version of FOIA) procedures underway to get insight in the agreement.

Albeit I am not a lawyer, a few thins caught my eye. The agreement is very similar to the agreement between Google and the British Library. Bert Zeeman pondered the idea of standard Google contracts in this respect. This seems to go for the exception of the number of volumes in the public domain that will be digitized, 250,000 in the UK and 160,000 in the Netherlands (clause 2.1).

What struck me as interesting was the use of the libraries digital copies, clause 4.8 “the library may provide all or any portion of the library digital copy… to (a) academic institutions or research or public libraries, ….” But we are not able to “providing search or hosting services substantially similar to those provided by Google, including but not limited to those services substantially similar to Google book search”. I guess that leaves out the other academic libraries in the Netherlands to include these digital copies in their discovery tools. It is tempting, but I see problems on the horizon. We seem to be left with separate information silos whereas integration with the rest of the collection would be really interesting. It becomes more explicit in clause 4.9 where it is stated that “nothing in this agreement restricts the library from allowing Europeana to crawl the standard metadata of the digital copies provided to library by Google.” We would be more interested in the data rather than the metadata.

But then again, it is up to the lawyers to see what’s allowed and what’s not. But then again, again, after fifteen years all restrictions on the use or distribution terminate (clause 4.7), a bit long according to the open rights group. However, we have experience with building academic library collections, it takes ages. Those fifteen years are over in the wink of a young girl’s eye.

Linking from Catalog of Wageningen UR Library to Google Books

Previously I announced that we made use of the Google Books API to link to the full text whenever possible. We only experienced two problems with this service. First, the quite frequent Google spam warnings, which have been partially resolved but still keep coming back. Second, we did not have the required OCLC or LCCN numbers for the pre-ISBN books in our catalog.

Thanks to OCLC Nederland this problem has been circumvented successfully as well. OCLC build a service which we feed the PPN (Pica Production Number) , which is available in our catalog and returns the OCLC number. We use that number to feed it into the Google Books API which determines the kind of electronic availability of those books which results in the right link and text on the catalog record. Peter described this in more detail. Just another Hooray for OCLC, since the service is now working.

A few examples are:

Even when the full text is not available on Google Books, the service can be usefull. In the following example of Hogg, R. (1884) The fruit manual, the electronic version of the 1860 edition is available on Google Books rather than the 1884 edition we have in our collection.

It took actually quite some effort to find these examples. Perhaps an indication of our unique collection?

Implementing the Google Book Search API at Library Wageningen UR

Last Thursday Google announced the Book Search API officially. That created some excitement in the Dutch Library world as well. The public library at Delft -those DOK boys from the Shanachie Tour– has implemented it in their catalog using the static linking options to Google Books.

Our application developers were also very interested in this new toy and made a first step of implementation for the books in our catalog based on the ISBN numbers available. Where we had Amazon book covers already available, kept these (eg. The genetic diversity of cacao and its utilization), on many occasions we already linked to the fulltext of books but could now include the Google Book cover to the catalog as well (eg: Return to resistance : breeding crops to reduce pesticide dependence). However the most interesting cases are of course those books available at the library, but which have now fulltext links through Google Books as well (eg: Illustrated guide to integrated pest management in rice in tropical Asia).

We have not reached the full potential of fulltext linking to Google Books yet. Our current implementation is based on ISBN only. So all books before 1965 have not yet been linked to Google Books. Our problem is quite simple, in the Netherlands we are not using OCLC numbers in the Central Dutch Catalog, albeit PICA is currently 100% part of OCLC. We are using Pica Production Numbers instead. So we have inquired in Leiden (OCLC the Netherlands), and they are inquiring in Dublin (OH) to get at PPN to OCLC conversion table or whatever. After we have resolved this little problem we van continue to link the older books as well.

Another problem we encountered is that the GB API results into spam warnings quite quickly and requires you to fill out a captcha. With the GB API implemented at in our catalog we run at the university quite quickly into problems. All traffic from the university has the same IP address, that of the firewall, which Google identifies quite quickly as illegal activity. I suspect other universities will experience similar problems. At home however, it works swell.