Identifying Discourse Continuities in “Unrelated” Ancient Chinese Texts with LLM
利用大語言模型識別中國古代「無關」文本中的話語連續性
Prof. Yegor GREBNEV
北師香港浸會大學副教授
Associate Professor, Beijing Normal-Hong Kong Baptist University
分類對於認識與理解古代文本是非常重要的。然而，我們習以為常的分類方式往往並不完善，甚至可能是錯誤的。當我們將文本歸類為「儒家」或「道家」等標籤時，有時反而會使我們難以追溯中國古代文化不同部分之間的重要聯繫。在本次講座中，我將說明如何利用大型語言模型來識別那些通常被視為彼此無關之文本之間的這些連續性。
Categories are essential for knowing and understanding ancient texts. Nevertheless, the categories that we are used to are often imperfect and even wrong. By classifying texts under such labels as “Confucian” or “Daoist”, we sometimes make it difficult to trace important connections between different parts of ancient Chinese culture. In this talk, I explain how LLMs can be used to identify such continuities in texts usually seen as unrelated.
Dr Yegor Grebnev holds a PhD degree in Oriental Studies from Oxford and a M.A. in Chinese History from Moscow State University. He has previously worked as a Junior Fellow at the Society of Fellows in the Liberal Arts at SUSTech in Shenzhen and as a Junior Research Fellow at Merton College, Oxford.
Dr Grebnev studies ancient Chinese history and texts. He is particularly interested in comparative research, using evidence from other ancient cultures to elucidate China’s past. Apart from his work on China, Yegor has initiated several international collaborations in the field of manuscript and epigraphic studies that bring together researchers working on a broad range of pre-modern cultures.