Since its discovery in 1912, the Voynich manuscript has proved to be a challenge for historians and cryptographers alike. Carbon dated to the 15th century, the manuscript has yet to be deciphered, but progress is being made as a team of scientists from the University of Alberta announced that applying AI-based algorithms has proved fruitful.
Could Machine Learning Decipher the Mysterious Voynich Manuscript?
A team of researchers from Canada’s University of Alberta might be on the right track to discover the meaning of the seemingly ambiguous writings contained by the 15thcentury Voynich manuscript. Using AI-based cryptography algorithms, the scientists started by addressing the manuscript’s language of origin.
Although many have speculated that the scroll might have been written in ancient Arabic, the AI analysis revealed that it’s actually ancient Hebrew. Greg Kondrak, a University of Alberta professor and one of the members involved in the recent attempt to decrypt the Voynich manuscript, declared that the AI cross-referenced the writing on the 240-page long manuscript with the Universal Declaration of Human Rights, which contains a database of 400 languages.
Upon feeding in the data into the AI, Kondrak and his colleagues discovered that approximately 80 percent of the words were in ancient Hebrew. To test out this new assumption, the researchers called in ancient Hebrew specialists and also tried running the processed text through Google Translate.
Baffling enough, the sentence processed by Google Translate was grammatically sound and could be further interpreted. It might not be much, the professor added, but this is definitely a great starting point.
According to Kondrak, the sentence that came out of Google Translate read:
She made recommendations to the priest, man of the house and me and the people.
At the moment, the team is working on improving the algorithm’s recognition rate and perhaps apply it to other ancient manuscripts.
Image source: Wikipedia