Research: Norbu Ketaka: A Neural Spelling Correction Model Built on Google OCR-ed Tibetan Manuscripts
Leonard W.J. van der Kuijp, Professor of Tibetan and Himalayan Studies, Department of South Asian studies and East Asian Languages and Civilizations; Chair, Committee on Inner Asian and Altaic Studies
While many researchers can now digitally access ancient Tibetan texts, spelling errors are still hindering their ability to generate reliable observations. Such errors stem from issues like slight differences in letters that are indistinguishable, even to the naked eye. For this research grant, ancient times and AI collide: Leonard W.J. van der Kuijp seeks to use artificial intelligence to create a system that can auto-correct spelling mistakes in current e-texts, allowing scholars to draw more effective conclusions. To help with his project, W.J. van der Kuijp will work together with a team of Harvard affiliates and collaborate with Tibetan experts and human annotators in China. Queenie Luo, a Harvard Ph.D. Candidate leveraging computational techniques to study Chinese society, will be responsible for developing the AI system.
Research: The Evolution of Ant Biodiversity in China’s Hengduan Mountains
Naomi Pierce, Sidney A. and John H. Hessel Professor of Biology, Curator of Lepidoptera in the Museum of Comparative Zoology; Senior Fellow of the Society of Fellows
Cong Liu, Associate, Department of Organismic and Evolutionary Biology
In the Hengduan Mountains, tiny ants hold the secret to a key part of our global ecosystem. As one of the largest biodiversity hotspots in the world, the mountains in Southwest China’s Yunnan province have explained a lot about the region’s plant species. However, our understanding of its insect population is extremely limited. Using ants as a case study, researchers have found that Hengduan’s ants are just as diverse as its plants, yet the underlying evolutionary processes are still unknown. To bridge this gap, Naomi Pierce and Cong Liu seek to further investigate mechanisms of how ants—and ultimately insects more generally—adapt, diversify, and form new species in the Hengduan mountains. In doing so, the team will collect and characterize ant specimens, and investigate them for evolutionary processes and thermal adaptations. To help with this endeavor, Pierce and Liu will collaborate with Professor Yan-Qiong Peng from the Chinese Academy of Sciences and Professor Guo-Jie Zhang from the Kunming Institute of Zoology.
Conference: Approaches to Cultural Heritage and Digital Media Art in the Modern Era
Eugene Wang, Abby Aldrich Rockefeller Professor of Asian Art, Department of History of Art and Architecture; Founding Director, Harvard FAS CAMLab
Scholars have yet to explore how cutting-edge technology can help people visualize data related to traditional Chinese heritage. Eugene Wang is planning a conference to be held at the Harvard Center Shanghai that will convene experts from many disciplines to discuss a new era of data usage for the cultural heritage field. The meeting will highlight the intersection of historical art, cultural preservation, and the role of digital media. Focusing on these dimensions will help scholars establish innovative ways to teach the public about China’s cultural heritage, and ultimately, reach a wider audience. Harvard CAMLab, a museum that showcases ancient culture through immersive and modern art installations, stands out as a successful example. In May 2023, CAMLab used over 5,000 data points, machine learning, and immersive technology for a cave dance installation. Buddhist apsaras were brought to life, as guests were surrounded by digitized dancers twirling and leaping across the exhibit.
For an update on Eugene Wang’s grant, please read, “Experiencing Buddhist Transcendence Through Technology.”