基于古籍NER语义分析的城市历史空间解译研究OA

The current urban historical space commonly faces the challenge of superficial cognition.The spatial interpretation methods based on the data of the relevant ancient books suffer from limitations such as low efficiency and difficulty in systematically extracting spatial information.To address this,this paper constructs an intelligent spatial interpretation method system for the ancient books based on named entity recognition(NER)and geographic information encoding.Taking the Annual Customs and Festival in Peking,a Qing dynasty folk customs chronicle,as the research subject,the paper adopts the techniques including text cleaning,named entity recognition,ancient-modern place name mapping,and geocoding,etc.to systematically extract 102 spatial entities and their associated folk activities in Beijing in the Qing dynasty,and constructs a three-dimensional data matrix of"time-space-activity",which is then mapped onto the modern urban space through a geocoding system.The study reveals the operational patterns in the urban space of Beijing in Qing Dynasty,including the seasonal activity cycles,hierarchical functional distribution,and spatio-temporal coupling.It provides a new technical paradigm and practical reference for the digital interpretation of the ancient books and the revitalization of the urban historical spaces.