How AI is creating a 1,000 year-old social network

The Spark

5th July 2017

The Spark newsletter is our fortnightly roundup of the latest innovation trends. To sign-up, click the register link in the footer of this page.

History and machine learning are not seemingly natural bedfellows. However, a new project in Venice, The Venice Time Machine is changing perceptions.

Computer scientist, Frédéric Kaplan will capture over 1,000 years of records from the city of Venice in a dynamic digital form, including maps, monographs, manuscripts and sheet music. The aim is to not only to open-up reams of hidden history to scholars, but also to enable the researchers to search, link and cross-reference the information, thanks to advances in machine-learning technologies.

Books will be automatically scanned, using a robotic arm to turn pages, algorithms will tease writing out of the scanned pages, turning them into digital text, and individuals appearing in the same document will form the nucleus of a social network, which can be linked to maps, property records and trade ledgers.

If it succeeds, it will pave the way for a more ambitious project to link similar time machines in Europe’s historic centres of culture and commerce, revealing in unprecedented detail how social networks, trade and knowledge have developed over centuries across the continent.

It would serve as a Google and Facebook for generations long passed.


Click here to read the more of this issue of The Spark.