Planned intervention: On Wednesday April 3rd 05:30 UTC Zenodo will be unavailable for up to 2-10 minutes to perform a storage cluster upgrade.
Published June 25, 2015 | Version v1
Poster Open

3. The Quality and Visibility of Library Digital Collections as Cornerstones of Open Science in Humanities: Belgrade University Library's Historical Newspaper Collection

Description

The poster presents work carried out by Belgrade University Library in creating a quality digital collection of historical newspapers and promoting it to users.

In the course of the Europeana Newspapers project (www.europeana-newspapers.eu) Belgrade University Library refined its digital collection of historical newspapers and acquired fully searchable METS/ALTO files for the entire collection. The collection consists of 46 publications printed on 400,000 pages. Although the OCR process was implemented with a high accuracy rate of 98%, the main complaint from researchers was about this one point. The less than perfect accuracy of OCR impeded both the precision of searching and the overall usability of the collection.

User demand is for 100% accuracy of the newspapers texts and to achieve this, a corrector for the METS/ALTO files was needed. Limited budgets made it necessary to develop an in-house solution which was created in 2014. Several innovative programming solutions which made this unique corrector possible will be presented in the poster. Once a functioning tool was available, there was then the problem of correcting thousands of pages. The solution so far is on-site work by volunteers, librarians and researchers willing to contribute their efforts in correcting texts. Real live data on the time needed for correction depending on publication date and content will be presented, along with the quality control process employed. The most up-to-date data at the point of finalisation of the poster will be provided.

The other important cornerstone making such a digital collection a true foundation enabling the principles of Open Science for researchers in humanities is effective visibility of the collection for researchers, and for the general public interested in participating in Citizen Science projects based on such collections. The poster will demonstrate the partnership between Belgrade University Library and the Serbian national broadcasting service in bringing information about this collection to the general public and to the specific user group of humanities researchers. About a dozen TV and radio shows have used historical newspapers from this collection in order to show best practice in its use, and a widely disseminated national web portal presented over 20 texts with the same aim over six months in 2014 and 2015.

Files