Dr. Tarin Clanuwat
Dr. Tarin Clanuwat is a senior research scientist at Google Brain based in Tokyo. She received her PhD in classical Japanese literature from the Graduate school of Letters Arts and Sciences at Waseda University, where she specialized in Kamakura-era Tale of Genji commentaries. In 2018, while she was a project assistant professor at ROIS-DS Center for Open Data in the Humanities (CODH), she developed an AI-based kuzushiji recognition model called KuroNet. In 2019, she hosted a Kaggle machine learning competition for kuzushiji recognition which attracted over 300 machine learning researchers and engineers from around the world. In order to make Japanese historical documents more accessible to the general public, Tarin also developed “miwo” AI Kuzushiji recognition smartphone application.
Her kuzushiji recognition research and applications won several awards including Yamashita Memorial Research Award from the Information Processing Society of Japan, research paper award from Japan Society for Digital Archive Conference and Excellent award from Vitalizing Local Economy Organization by Open Data & Big Data.
"The journey to make billions of Japanese historical documents accessible"
Japan is a unique country with one of the world's most distinctive cultures. Over a thousand years of Japanese culture and knowledge are preserved inside historical documents, currently inaccessible to 99.99% of Japanese people to whom the documents are unreadable. They are written in Kuzushiji, or cursive Japanese, which looks like patterns on paper for most people. Hence, in order to preserve the culture of Japan, collaboration and participation of the general public is indispensable. In this keynote, I will describe what Kuzushiji is, what makes it so important and how machine learning can help us unlock and shed some light on our rich history. I will also talk about the efforts of many organizations in tackling this problem and how we collaborate despite completely different approaches to the same goal: to make Japanese historical documents accessible to everyone.
dr. Jieh Hsiang
Jieh Hsiang is a Distinguished Professor in Computer Science of the National Taiwan University. He is also the director of the NTU Research Center for Digital Humanities, the first such center in the Sinophone world. In his pre-DH life, he received a Ph.D. in computer science from the University of Illinois at Urbana-Champaign and worked mainly in automated theorem proving, in particular term rewriting systems. After returning to Taiwan in 1993, Jieh Hsiang started to work with with historians and anthropologists and initialized the digitization of Taiwanese cultural heritage at his university. Through the years, he and his team built over 30 large scale digital libraries of Chinese/Taiwanese historical archives, all of which utilize a context-discovery retrieval methodology that is designed for scholarly use of digital archives. His team also developed DocuSky, a personal DH platform for humanities scholars to process, annotate, analyze and visualize data and build their own searchable text databases without the help of an IT specialist. The NTU Digital Library of Buddhist Studies, a comprehensive bibliographic on-line database that he has been in charge of since 2005, attracts more than 11,000 users around the world each day.
Jieh Hsiang was the author/editor of six books on digital humanities in Chinese. Being the first of its kind, the books made significant impact in promoting DH in the Sinophone world. He organized the first Conference on Digital Archives and Digital Humanities in 2009, which becomes the first annual DH conference in East Asia. He also helped establishing the Taiwanese Association for Digital Humanities in 2016, and the first Chinese language journal on DH, the Journal of Digital Archives and Digital Humanities, which was inaugurated in 2018. Jieh Hsiang served 6 years as the University Librarian of NTU and 8 years as the Director of NTU Press. Before returning to Taiwan, he was a full professor in Computer Science at Stony Brook University.
"Harnessing Chinese Historical Writings for Digital Humanities "
About nine hundred years ago Zheng Qiao (鄭樵, 1104~1162 CE), a Chinese scholar of the Southern Song Dynasty, ventured on an ambitious project of “collecting all books into one book”. The backdrop of his motivation was that China had a uniform written language for more than a millennium at the time (more than two at present) with a long tradition of keeping historical records even with the dynastic changes. To illustrate his idea, Zheng Qiao wrote a book called Tongzhi (通志, Comprehensive Treatises) in which he condensed pre-Tang Chinese history into 200 volumes (about 5,000,000 characters). For example, in the Outline of Anomalies (《通志。災祥略》), he listed all the anomalies such as solar eclipses and earthquakes that he could find from historical records and used their frequencies to argue that, contrary to common belief, such natural phenomena did not correlate to how benign the ruler was (and therefore were not warnings from heaven). Zheng Qiao was way ahead of his time and Tongzhi had largely been ignored by later scholars. The emergence of digital humanities has cast new lights on Zheng Qiao’s vision.
In this talk, we shall give a brief overview of the issues and challenges facing the digitization of Chinese historical writings, together with the current state of digitally available searchable full-text. We shall describe the importance of tagging (of time, person, place, and objects), how they are done, and how they can be used to connect documents to create textual contexts. We shall present a theory of (textual) context discovery and show how it can be used to build retrieval systems aimed at discovering, analyzing, and visualizing contexts hidden among documents. A platform, DocuSky, which enables humanities scholars to build their own personal context discovery retrieval systems without the help of IT specialists will also be presented. We shall complete the presentation by giving a short discussion of how Zheng Qiao might have done were he living in the DH era.
the winner of the triennial Antonio Zampolli Prize:
Voyant Tools / Dr. Geoffrey Rockwell
Dr. Geoffrey Rockwell is a Professor of Philosophy and Digital Humanities at the University of Alberta, Canada. He is also the Director of the Kule Institute for Advanced Study and Associate Director of AI for Society signature area. He received his MA and PhD in Philosophy from the University of Toronto where he also worked in computing as a Senior Instructional Technology Specialist. From 1994 to 2008 he was at McMaster University where he directed the Humanities Media and Computing Centre and helped found the Department of Communication Studies and Multimedia.
Rockwell publishes on video games, textual visualization, text analysis, ethics of technology and on digital humanities including a co-edited book on Right Research: Modelling Sustainable Research Practices in the Anthropocene (Open Book Publishers, 2021) and a co-authored book Hermeneutica: Computer-Assisted Interpretation in the Humanities (MIT Press, 2016). He is co-developer of Voyant Tools (voyant-tools.org), an award winning suite of text analysis tools. He is a co-organizer of the Replaying Japan conference, co-editor of the Journal of Replaying Japan, and President of the Canadian Society for Digital Humanities.
Voyant Tools is an in-browser suite of tools for the study of electronic texts. You upload a text to the Voyant server or use one already indexed and then you have access to coordinated set of analysis and visualization tools. Voyant will show you trends over the text or allow you to compare different texts in a corpus. It will provide textual statistics and word lists that you can export for use in other tools.
Voyant was designed by humanists for the interpretation of texts. It doesn’t give you answers, it allows you to explore small or large text collections intuitively. As it is free and runs in the browser on our server it is used for introducing students to text analysis. To learn more about the computer-assisted interpretation of texts with Voyant see our book:
Rockwell G. & S. Sinclair. Hermeneutica. MIT Press, 2016
Voyant Tools is at http://voyant-tools.org
We have hands-on tutorials at http://dialogi.ca
A downloadable standalone version is available: https://github.com/voyanttools/VoyantServer