The Vatican Secret Archives comprise 600 collections of texts spanning 12 hundred , most of which are nearly impossible to access . The Atlanticreports that a team of scientist is hoping to convert that with help from some high school students and artificial intelligence software .
In Codice Ratiois a new research project dedicated to analyzing the vast majority of Vatican ms that have never been digitalise . When other libraries wish well to make a digital archive of their inventory , they often use optical - case - credit ( OCR ) software package . Such program can be trained to recognize the letters in a sealed first rudiment , pick them out of hard - transcript ms , and convert them to searchable text . This technology pose a challenge for the Vatican , however : The many older text in its collections are written by hired man in a cursive - corresponding script . With no space between the character , it ’s unacceptable for OCR to determine what ’s a letter and what is n’t .
To get around this , the inquiry team at In Codice Radio tweaked OCR package so that it could recognise pen stroke instead of letters . The OCR can discover the pen stroke that make up letters in an alphabet by looking for spots in the text where the ink narrows rather than portray full gap between characters . The strokes are n’t very useful on their own , but the software can conflate the piece to shape possible letters .
To serve the software do even well , research worker recruited students from 24 Italian high-pitched school to check over its work . As the researchers explain intheir newspaper , the students were shown a list of acceptable edition of a real letter , such as the letter A , and were then throw a list of characters the software system had guessed might be the real letter . By selecting the characters that match the acceptable versions , they were able to tardily instruct the software the mediaeval Latin ABC .
All this information , plus a database of 1.5 million Romance intelligence that had already been digitized , eventually brought the OCR to a place where it could use artificial intelligence to identify real letter on its own . The final result are n’t gross — a upright portion of the Word transcribed so far hold typos — but Vatican archivist are a lot better off than they were before : The computer software can identify single handwritten letters with 96 percentage truth , and misspelled words can still ply important circumstance to readers . The goal is to eventually use the software to digitalise every document in the Vatican Secret Archives .
[ h / tThe Atlantic ]
