National Library of Australia Staff Papers, 2009

Font Size:  Small  Medium  Large

Many Hands Make Light Work: Public Collaborative OCR Text Correction in Australian Historic Newspapers

Rose Holley

Abstract


The ANDP team had from the outset in January 2007 decided to make a considerable investment in software development in order to be able to quality assure digital outputs to ensure they met minimum standards and to future proof the files in case they could be improved further in the future.

Contributors had all expressed concern that the digital outputs (image quality, OCR text) may not be good enough to enable adequate full text retrieval or to meet user expectations. The ANDP team had regularly brainstormed and reviewed ideas to improve the quality of outputs and had implemented a number of ideas to achieve this3. At this stage the team were assuming that quality of data entirely relied on the Library’s digitisation technique. It had not yet occurred to the team or the contributors that the public may play a role in improving and enhancing the quality of the data.

Full Text: PDF