The proposed project is the first year of research on a three-year term basis. The whole project focuses on providing an alternative to the stereotypes in the worldview of Han Chinese, restoring the historical reality of the frequent contacts between different minorities and the Han state during the Wei-Jin Southern and Northern Dynasties, deconstructing the prejudice in Chinese tradition towards the “strange things from the marginal regions”. By using several special topics, it intends to construct a more diverse image of the periphery, and analyze the structure of feelings expressed about it.
By means of Text Mining and markup of authority terms, during the first year it will set up the authority terms, collect related material, and investigate foreign names, which it will connect to geographical distribution, historical power and social order. More specifically, the project will start with digital editing of lost textual fragments based on the work done by the research team in recent years in the sphere of event extraction from Chinese medieval texts. Digital editing is founded on Term Extraction and includes the pre-training model of Transfer Learning, the calculation of semantic characteristics of Self-Attention, and matches the Text Mining, which will be done by Next Sentence Interference in the Masked Language Model. The project will use the narrative of peripheral/foreign regions in accounts of foreign countries, official histories, gazetteer records, short stories, Buddhist monk biographies, and travel records to foreign countries from the Wei-Jin Southern and Northern Dynasties. It will rely on the Academia Sinica Digital Humanities Research Platform and its digital tools to manage the texts, perform the markup, and construct the thematic hierarchy of authority terms. To discover the condensed image of the periphery, the project will apply the methods of mining of digital texts and markup of attributes of authority terms, as well as statistics of word vectors, word quantification, and word co-occurrence. Based on the general idea of digital compilation of the periphery narrative, the first year will focus particularly on the extraction, statistics, and analysis of foreign names, unveiling the important function names have in consanguinity, as well as in diverse geographical, political and matrimonial senses, thus establishing a quantitative index on group density and deepening the discourse on this topic.