AxisMundi Jerusalem

Sorting through a digital history

With more and more libraries digitising their archives, academics have a growing number of texts they can access without having to get on a plane and journey to distant continents. Perhaps in the near future, researchers will be able to simply log on from their office to view a database of a nearly infinite number of ancient texts, prayers or whatever writings have been handed down by our ancestors.

Of course, problems arise with digitising thousands of years of handwritten documents. Making a digital copy is the easy part. Helping the computer understand what is written, well, that is a tough one.

Gideon Ben-Zvi, who has founded a couple companies in the field of Optical Character Recognition (OCR), told me that: “The eyes outperform even the best OCR software by magnitude, although the speed achieved by OCR is far faster than humans.”

That means a researcher can sit for hours in front of a page of text and will always emerge with a better understanding of the words written. However, once a computer program can discern words, phrases and even handwriting in the most highly degraded texts, you can then search through millions of pages almost instantaneously.