Research on the structuring method of text data in the field of power generation
Authors: Lan Mengqi, Huang Luyao, Wei Chen, Fu Junwei, Yu Jiawen, Yang Xiaorong
（Zheneng Shaoxing Binhai Thermal Power Co., Ltd）
Abstract: The rapid growth of the Internet has necessitated the use of intelligent and automated methods for processing vast amounts of power equipment information by professionals in the power generation sector. However, existing annotation tools often struggle to support the complexities of technical and specialized texts. This highlights the need for effective data annotation in the construction of knowledge graphs, particularly for optimizing the management of power plant equipment failure data. This paper presents a novel text data structuring method based on deep learning, which integrates multiple algorithms to optimize the text labeling process and achieve precise corpus annotation. The method demonstrates its effectiveness in building knowledge graphs within the power generation industry, leading to improved management of power plant equipment failure data. The analysis of real-world power plant equipment failure data verifies the potential of the proposed text data structuring method for establishing a visual analysis system for power plant fault diagnosis, ultimately enhancing equipment management capabilities.
Keywords: corpus annotation; Data structuring; Power equipment; knowledge graph
