Recently, Digital Humanities Laboratory of Shandong University successfully developed an intelligent assistant based on the borderland large model named“Chinese Borderland Intelligence Nexus (CBIN)”. Through core technologies such as multimodal parsing and knowledge graph construction, CBIN has achieved full-process intelligent processing of borderland historical documents from digitization to knowledge transformation, providing technical support for the construction of an independent knowledge system for China's borderland with Chinese characteristics.
Breakthroughs in Key Technologies. Professor Miao Wei, the project leader, introduced that CBIN has achieved three major innovative breakthroughs: it adopts a localized deployment plan to ensure the security and controllability of cross-border ethnic languages and sensitive historical data; it has established a "Frontier Historical Dataset" containing millions of documents and a "Frontier Spatiotemporal Knowledge Graph" with 200,000 entities; it has realized functions such as automatic punctuation of ancient texts and multilingual entity extraction, improving processing efficiency by 30 times compared to traditional methods and achieving an entity recognition accuracy rate of 96.3%.
Empowering Academic Research Innovation. Currently,CBIN has been applied to multiple national projects of the Laboratory. It has enabled the intelligent iteration of the original East Asian Digital Humanities Platform and the intelligent transformation of the literature database; it has empowered the intelligent organization and knowledge mining of ancient texts, old maps, and epigraphic documents required by a number of ongoing projects, greatly improving research efficiency; it is currently preparing to establish a "Parallel Corpus of East Asian Classics on Borderland in Chinese"， providing new data support for strengthening the construction of a community with a shared future among China and its neighboring countries, enhancing cultural integration and strategic mutual trust.
Building an Open Research Ecosystem. Professor Miao Wei emphasized that CBIN is not only a technical tool but also a bridge connecting history and reality, academia and society. Traditional borderland research relies on scholars to manually organize fragmented information, while CBIN can "translate" documents into computable data, allowing researchers to focus on higher-dimensional knowledge discovery. In the future, we will open API interfaces to jointly build a digital and intelligent ecosystem with the academic community. The Laboratory plans to open-source some data and intelligent applications based on CBIN by the end of 2025 and launch the "Borderland Intelligent Research Collaborative Innovation Initiative"together with Zhipu AI, Heywhale, and some universities.
Source: Shanda View https://www.view.sdu.edu.cn/info/1021/201001.htm
