Minna de Honkoku: Citizen-Participation Transcription Project for Japanese Historical Documents

Abstract

Minna de Honkoku began as an online citizen science project to transcribe earthquake-related historical materials from the Earthquake Research Institute Library of the University of Tokyo. In Japan, almost all the documents are written in kuzushiji (old-style Japanese cursive script), a writing style used before ∼1900. Because the style of writing is different modern Japanese, transcription is necessary to use the historical documents as data for earthquake research. The workspace of Minna de Honkoku consists of a viewer of a document image and a vertical (Japanese-style) editor for transcription. Users can input transcribed text while viewing its image. The ranking of characters transcribed is displayed to keep users motivated. As of October 2024, more than 9,700 people were registered for the project, with the total number of characters transcribed at about 41 million. The text generated by Minna de Honkoku can be used for various academic research fields including seismology and can be used to enhance citizens’ disaster awareness. The paired kuzushiji characters and text data generated by Minna de Honkoku are beginning to be used as training data for artificial intelligence. ▪ Minna de Honkoku is an online citizen science project aimed at deciphering historical documents. ▪ The total number of participants is 9,700, and characters transcribed by Minna de Honkoku reaches 41 million. ▪ Minna de Honkoku began as a project to transcribe earthquake-related historical materials. ▪ The text generated by Minna de Honkoku is used in seismology and various research fields and for building artificial intelligence–based kuzushiji recognition.