Research On Knowledge Graph Construction Method For Power Communication Planning Field

The power communication network is an important infrastructure of the power system,and there is a large amount of knowledge in it.The use of this knowledge can provide a guarantee for the safe,stable and reliable operation of the power system.Power communication planning text,as a form of knowledge representation of power communication network,has the problem of difficulty in acquiring knowledge.With the research in recent years,knowledge graph technology can carry out a perfect structured representation of knowledge,and has been successfully applied in various fields.At present,there are few studies on the construction of knowledge graphs in the field of power communication networks,and the relevant knowledge bases have not been constructed yet.In order to make full use of the knowledge of power communication planning,this paper studies a knowledge graph construction method for the field of power communication planning.By analyzing the research and application of knowledge graph at home and abroad and the principles of related deep learning models,this paper studies the graph construction technology based on deep learning,and proposes a knowledge graph construction framework for power communication planning.In the key knowledge extraction part,a named entity recognition model based on joint word embedding and Transformer is proposed,and a segmented convolution network relation extraction model based on hole convolution is proposed.The input form of word vector is improved by joint word embedding,so that it contains richer semantic information.The Transformer structure is used to improve the original recurrent neural network series entity recognition model,so that it can perform parallel operations and can extract deeper features.The problem of insufficient receptive field of segmented convolutional neural network is improved by atrous convolution,and more contextual semantic information is associated.Finally,the extracted knowledge is visualized using the Neo4j graph database.The improved model in this paper is verified by simulation and adopts a unified evaluation standard,and the performance has been improved on the basis of the original model.The constructed knowledge graph can clearly display each entity and the relationship between entities through visualization,which shows the effectiveness of the model and method proposed in this paper.It reflects the knowledge representation ability of knowledge graph in the field of power communication planning.