Construction of Knowledge Graph of Historical and Cultural Towns and Villages in Anhui Province

Since 2005, the number of historical and cultural towns and villages has been decreasing by more than two hundred per year. At present, the research on historical and cultural towns and villages mainly gathers in the field of village spatial morphology and geography, while there is a lack of research in the field of digital development of famous towns and villages. This dissertation proposes to apply knowledge graph technology to the field of historical and cultural towns and villages in Anhui Province, in order to solve the problems of low relevance and low query efficiency of the related knowledge of the famous towns and villages; at the same time, we use Bert-BiLSTM-CRF based on adversarial training as a named entity recognition model to recognize the entities of the unstructured text of the related knowledge of the historical and cultural towns and villages in Anhui Province, and match the extracted entity pairs with a relational dictionary to get the corresponding entity pairs. Match the extracted entity pairs with the relation dictionary to get the corresponding relations. Combine the extracted and matched triples with Py2neo and Cypher statements to realize the automated construction of the knowledge graph of famous historical and cultural towns and villages in Anhui Province.