One of my favorite digitization projects to talk about (as a sort of “work of wonder”) is the digitization project to reconstruct thousands of shredded documents that were once housed at the East German secret police (Stasi) headquarters in Berlin.
Activists were able to save some 15,000 sacks of shredded paper from being destroyed in 1989. Now, however, archivists have to try and put the shredded pieces of paper back together – this is no easy task! It has taken nearly 20 years to create a computer system capable of accomplishing this task, but it looks as if progress is finally being seen.
The archivists use scanners to scan in all of the bits of paper and the computer attempts to piece them back together – like one huge puzzle. The computer is trained to look not color variations that could or should go together, but also analyse the edges of the paper to see if they physically can be placed next to each other. Unfortunately, the computer program is still having difficulty putting together words, sentences and content. This might seem strange, given the fact that text recognition has been so well developed already. However, the Stasi developed its own manner of writing reports that is difficult for outsiders to decipher, with many references to code words, assets and operations that do not make sense out of that spy-world context.
Developing this complex process has not been cheap. The German federal government has already invested about €6 million (approximately $7.9 million). To develop the technology, the federal government turned to the Fraunhofer Institute, the same people who brought you the MP3 by the way. The initial trial period is supposed to run just under two years and attempt to recover about 400 of the sacks of shredded documents. If the Fraunhofer Institute can demonstrate success, it is expected that the federal government will then contribute additional funds to complete the digitization process.
Here’s a five-minute video from Deutsche Welle TV (in English) that highlights the progress that they have made thus far. It is really worth watching!