Harvard puts metadata for 12M library items into the public domain

"The records consist of information describing works—including creator, title, publisher, date, language, and subject headings—as well as other descriptors usually invisible to end users, such as the equalization system used in a recording. "<p>I'm having a hard time thinking of what could be done with this data besides a library catalog.

Why not link directly to the official press release: <a href="http://isites.harvard.edu/icb/icb.do?keyword=k77982&#38;pageid=icb.page498373" rel="nofollow">http://isites.harvard.edu/icb/icb.do?keyword=k77982&#38;page...</a>

It's about time they did this. Harvard's was about the only major library that didn't allow Z39.50 access to their full MARC21 records. As a private individual with a large rare and antiquarian book collection, I welcome the news, since I've found that Harvard sometimes has the only other copy of a book I'm cataloging. A few other libraries require you to jump through some hoops to get to the data (British National Library, for example), but Harvard was shutting everyone other than faculty and alumni out.

I hate that the article title says "Big data for Books".<p>Here's a hint on how you can get a sense of whether you are dealing with "Big Data": <i>IF I CAN FIT IT ON A THUMB DRIVE, IT ISN'T BIG DATA</i>.

The direct links for API access and download (3.16GB) is given in the DPLA Dev Blog: <a href="http://blogs.law.harvard.edu/dplatechdev/" rel="nofollow">http://blogs.law.harvard.edu/dplatechdev/</a>

It seems that bittorrent would be the logical choice for distributing the dataset. I wonder if this is an oversight or if they are not expecting many people to download the dataset...

I'm not sure who would need to actually fill out the submission form. But wouldn't this: <a href="http://aws.amazon.com/publicdatasets/#3" rel="nofollow">http://aws.amazon.com/publicdatasets/#3</a> be convenient for working with a data set like this?

Universities are doing some pretty cool stuff with data. Every tech uni is now getting their students to work on social media data analysis. More exciting than entity relationship diagrams...

It's going to be interesting to see what people build and/or analysis they do with this data.

&#62;Finally, note that Harvard asks that you respect community norms, including attributing the source of the metadata as appropriate.<p>That's not what "public domain" means. If they wanted attribution, there are licenses for that. "Public domain" means that it belongs to all of us now. In that case, attribution is meaningless.