How a literary scholar is using machine learning to help unlock Japan's past
With help from TensorFlow
Old scripts become extinct when they're no longer known or read, like the ancient Japanese script Kuzushiji.
Literary scholar Tarin Clanuwat is part of a small number of people who can understand the near-extinct script.
With billions of documents written in Kuzushiji, Tarin taught herself how to use TensorFlow, Google’s open-source machine-learning platform, to transcribe them into modern Japanese.
Fully customisable and available for researchers everywhere, this tool may help unlock rare Japanese history, science and culture dating back to the 8th century.
Thanks to Tarin's use of machine learning, she's made Kuzushiji even more accessible to people today.
📚
{[slideCounterCtrl.currentSlide + 1]} / {[slideCounterCtrl.slideCount]}