Azernews.Az

by Alimat Aliyeva

Researchers from Stockholm University have developed an artificial intelligence-based tool that can significantly accelerate the decoding of historical ciphers, AzerNEWS reports.

According to project leader Professor Beata Medjesha, the system was able to decipher the famous Borg Cipher from the 17th century in just 28 minutes, whereas this task previously required weeks of manual work by researchers.

The Borg Cipher is a 400-page manuscript containing secret medical prescriptions and pharmaceutical knowledge used for treating various diseases. For instance, one of the recipes suggests drinking several glasses of high-quality wine or infusing nutmeg in dough to treat intestinal disorders. Scholars believe the manuscript was likely created in Northern Germany or Scandinavia during the 17th century. The cipher itself consists of 34 different symbols, including special graphic characters and Latin letters with diacritical marks.

The new AI tool works by combining character recognition, pattern reconstruction, and language modeling to gradually reconstruct the original text. The database developed by the researchers already includes more than 20,000 encrypted historical documents and fragments.

Experts note that around 1% of historical manuscripts preserved in archives and libraries worldwide are still fully or partially encrypted. Decoding them could provide valuable insights into diplomacy, everyday life, medicine, and cultural practices of past centuries.

Interestingly, historians point out that such medical recipes often mix practical knowledge with superstition, reflecting how science and folklore coexisted in early modern Europe. Some ingredients mentioned in similar manuscripts—like herbs, minerals, or alcohol-based mixtures—are now being re-examined by researchers interested in the history of medicine.

At the same time, the scientists emphasize that artificial intelligence does not replace historians, linguists, or philologists. Instead, it serves as a powerful assistant, helping them analyze complex texts faster and uncover previously inaccessible sources.

During testing, the system was validated on already deciphered documents, including the Borg Cipher, to ensure accuracy and reliability before being applied to unknown materials.