Innovative digital technology is opening up new possibilities for the preservation and dissemination of traditional knowledge and cultural heritage. Promoting the synergy between technologies and the humanities, a research team at The Hong Kong Polytechnic University (PolyU) has leveraged artificial intelligence (AI) technologies to distil, represent and visualise geographic and ecological knowledge from ancient Chinese texts. Their approach has helped to overcome the limitations of traditional text studies, while fostering the digitisation of ancient texts to enable more innovative models of knowledge inheritance.
The pioneering project - "Knowledge Integration of the Classic of Mountains and Seas: Reconstructing Ancient Ecological and Geographical Knowledge Heritage with Artificial Intelligence" - is led by Prof. HUANG Chu-ren, Chair Professor of Applied Chinese Language Studies of the PolyU Department of Chinese and Bilingual Studies. The only winning endeavour from a tertiary institution based in Hong Kong and Macao this year, the project was recognised as one of the "Top 10 Innovative Exploratory Projects" in the Tanyuan Scheme 2024, among 79 submissions from 48 universities and research institutes. The Scheme was initiated by Tencent and various cultural units in the Mainland and guided by the Department of Science, Technology and Education of the National Cultural Heritage Administration.
As the first part of the pre-Qin classic text Classic of Mountains and Seas (Shanhaijing), the Classic of Mountains (Shanjing) comprises approximately 20,000 characters of geographic information relating to mountains and rivers, as well as the natural resources therein. It is regarded as China's earliest attested compendium of geo-ecological information. Existing studies of the text remain within the classical humanities paradigm. This research project has however proposed a knowledge integration approach to present an innovative solution for modernising and transmitting the knowledge content of the Classic of Mountains, with the aim of arousing greater research interest in the field.
The research team uses various AI technologies including information extraction, knowledge graphs, graph retrieval-augmented generation (RAG), and large language model (LLM), to analyse, organise, integrate, and present the geographical and ecological knowledge embedded in the text. Their goal is to establish a systematic knowledge platform that can serve as a blueprint for the development of diachronic geo-ecological knowledge systems of ancient China through future studies into other ancient texts and documentation from different dynasties. This will enable research into the geo-ecological changes and variations in China over the past two millennia and provide appropriate environmental information to better understand historical events. As a first step, the team is developing a Q&A platform that consolidates information about the landmarks in Classic of Mountains, together with a digital map that visually illustrates the geographical features described in the text.
Rendered demonstration for the digital map of Shanjing
Prof. Huang said, "A deep and systematic understanding of ancient geography and ecology provide us with longitudinal information about how our environment has changed. This is also crucial for contemporary ecological governance. The project represents an interdisciplinary effort to integrate digital technology in humanities studies to address the challenges of the fragmentation of ancient literature and the usability of traditional knowledge in today's evidence-based studies, which will ultimately enhance the transmission, integration and application of historical knowledge in contemporary contexts."
Prof. Huang is the first Chinese permanent member of the International Committee for Computational Linguistics. He has also been recognised as one of the World's Top 2% Most-cited Scientists 2024 by Stanford University in the field of AI with his long-standing expertise in computational linguistics and digital humanities. His team members include PhD student Ms Ke LIANG and postdoctoral researcher Dr Xuemei TANG from the Department of Chinese and Bilingual Studies. Dr Qi SU, Associate Professor from the Research Center for Digital Humanities at Peking University, and Dr Jinghang GU, Research Assistant Professor of the PolyU Department of Chinese and Bilingual Studies are co-principal investigators.
