Colleges collaborate on libraries digital storage facility

A new online storage facility for the digital conversion of library materials is an elephant, both in name and in size.

HathiTrust, which is named after the Hindi word for elephant, was launched last Monday in an effort to consolidate the location of books and journals digitally scanned and uploaded by member institutions of the Committee on Institutional Cooperation.

CIC Assistant Director for the Center for Library Initiatives Kim Armstrong said HathiTrust was the next step in the way CIC institutions collaborate with each other.

She said the CIC is looking at making things more available in an efficient manner.

“We’re literally pulling things off the shelves and making them digitally available,” Armstrong said.

Current members of HathiTrust include all 13 CIC institutions — of which the University of Minnesota is a member — as well as the University of California system and the University of Virginia.

Under the CIC-Google Agreement, Google has been digitally scanning millions of volumes from CIC libraries since 2007.

Digital Collections Unit head Jason Roy manages the daily digital scanning of books at the University. The unit has made 110,000 individual scans at the University over the past two years alone.

Google has already collected so much digital library material — well into the petabytes, more than 10,000 gigabytes, Roy said — that the CIC institutions, including repository administrators the University of Michigan and Indiana University, created HathiTrust to ensure that all those files were safely stored somewhere.

The University of Michigan hosts the main HathiTrust servers. Indiana University, meanwhile, is home to HathiTrust’s mirror servers which backup Michigan’s information.

Indiana Dean of University Libraries Patricia Steele said the main value of the CIC-Google project is Google’s vast amount of money and resources that allows them to efficiently preserve library materials.

“That said, they are a commercial entity,” Steele said. “They can be here for only 50 years; [institutional libraries] have been here for hundreds of years. We want to make sure we took the same care of the digital aspect of a resource as in print.”

Only titles that are in the public domain — currently about 16 percent of all HathiTrust content — will be available to students.

All HathiTrust’s content so far has been provided by Michigan and the University of Wisconsin, with materials from California, Virginia, Indiana and Purdue soon to follow.

For the time being, the materials can be accessed through Google or search engines on library websites.

Eventually, HathiTrust will have its own online search technology, as more materials become available through copyright permissions and the public domain.

“We felt that many of these titles are brittle or acidic, and just being able to have that digital file will be important,” Armstrong said.