Research And Design Of Knowledge Modeling And Knowledge Graph Construction For Sea Battlefield

With the development of naval battlefield information technology,military operations have put forward higher requirements for the storage and mining of naval battlefield knowledge.Naval battlefield data includes massive spatial data such as water surface,underwater,sky,and islands,and has many types of equipment,situation,events,etc.Knowledge graph technology makes up for the insufficiency of the difficulty of processing massive data,inflexible modeling methods,indirect expression and low data utilization rate in traditional databases.It can provide important knowledge basis for naval battlefield data in military command.However,the application of knowledge graphs in the field of naval battlefields is not mature,and a relatively complete naval battlefield knowledge system is in urgent need of construction.This paper takes knowledge modeling method and knowledge graph technology as the main technical points to construct a knowledge model and prototype system for naval battlefields.The main work of this paper is as follows:1)Propose a new knowledge modeling method--spiral feedback method based on demandIn order to solve the problem of building a knowledge model for naval battlefields,and aiming at the strong military demand of the knowledge system,a demand-based spiral feedback method combining the seven-step method and the spiral model was proposed,and completed the naval battle activities,naval battle equipment,military and political events.The construction of the ontology knowledge model of the naval battlefield domain based on core concepts such as prior knowledge of experts.2)Knowledge extraction for naval battlefieldsIn order to get rid of the dependence on artificially defined features and improve the efficiency of entity extraction,data preprocessing is based on the CBOW model and BIO rules,and the BiLSTM-CRF model is selected for the extraction of sea battlefield entities;for the problem of less open data on sea battlefields,based on dependency syntax analysis Construct a seed template for naval battlefields,and select Bootstrapping algorithm to achieve relation extraction.3)Realization and application of knowledge graph for naval battlefieldIn order to meet the needs of naval battlefield data management,the Neo4j graph database platform is used to realize the storage of entities and the relationships between entities,display the visualization and query effects of the knowledge graph.The knowledge graph for the naval battlefield performs well in intelligent search,knowledge question and answer,intelligent recommendation and explanation mechanism.