Study On The Construction Of Knowledge Map Of Red Literature Resources In The Perspective Of Digital Humanities

With the booming development of digital humanities research,it has not only brought great impact to the development of disciplinary fields such as history,art,and archaeology,but also brought unprecedented changes to the field of documentation and archives.In recent years,digital humanities has thrived in the Chinese academy.As one of the most important carriers of political memory in contemporary China,the digitization of red documentary and archival resources is conducive to the transformation from mere two-dimensionality to multi-level interaction,thus changing the mode of knowledge production and the form of results,realizing the cross-border integration of humanities research and science and technology,making full use of red documentary and archival resources,and injecting new vitality into humanities research.Based on a comprehensive analysis of the characteristics of red information resources and user needs,the research presented in this paper explores the key issues of building a knowledge graph of red information resources,which include modelling red information resource knowledge,acquisition of red information resource knowledge,storage and application of red information resource knowledge graph.The main research carried out is as follows.(1)Ontology construction of official documents of Shaanxi-Gansu-Ningxia Border Region Government.Since there is no mature ontology construction method in the field of red documents,this study is based on the application scenario of Shaanxi-Gansu-Ningxia Border Region government official documents resources.Firstly,we fully analyze the characteristics of the original data of Shaanxi-Gansu-Ningxia Border Region government official documents;select an ontology construction method suitable for the field of Shaanxi-Gansu-Ningxia Border Region government official documents;describe the ontology of ShaanxiGansu-Ningxia Border Region government official documents with ontology language,and then realize the ontology construction based on Protégé software,and present the concept hierarchy to establish the foundation for the subsequent construction of the knowledge graph of Shaanxi-Gansu-Ningxia Border Region government official documents.(2)Knowledge graph construction of official documents of Shaanxi-GansuNingxia border area government.The raw data in the official documents of Shaanxi-Gansu-Ningxia Border Region government mostly use unstructured data.In this paper,we firstly propose the BERT-Bi LSTM-CRF named entity recognition method based on deep learning algorithm,extract the entities of Shaanxi-GansuNingxia Border District government official documents,divide the hierarchical relationship categories and attribute categories according to the ontology layer,store them by using Neo4 j graph database,and finally complete the construction of the knowledge map of Shaanxi-Gansu-Ningxia Border District government official documents resources.(3)Knowledge discovery of official documents resources of Shaanxi-GansuNingxia Border Region Government based on knowledge graph.The content and topics of red literature resources are presented in the form of knowledge graphs based on correlated data through the Neo4 j graph database,and multi-dimensional analysis is conducted on the official documents of the Shaanxi-Gansu-Ningxia Border Region Government.With the help of query and inference methods,the knowledge discovery of the official documents of the Shaanxi-Gansu-Ningxia Border Region Government is completed,which extends and extends the Linked Data from the narrow sense to the broad sense,remedies the problem that the content attributes of the red documents are not fully revealed in the existing research,and proves the advantages of the graph database in the construction of the knowledge map of the red document resources.The deep mining of red literature resources brings the possibility of achieving accurate and intelligent services in the field of red literature resources in the future.