Power grid infrastructure engineering data knowledge graph construction method based on machine learning

Abstract

The invention discloses a machine learning-based power grid infrastructure project data knowledge graph construction method, and the method comprises the following steps: extracting information in unstructured text data and semi-structured table data of a power grid infrastructure project by utilizing a dependency relationship model and a bidirectional long-short-term memory artificial neural network hybrid model; mining the valuable information contained in texts in the data through knowledge fusion, and then constructing the capital construction project data knowledge graph. According to the graph, natural languages in different types of files are converted into nodes and relationships in a semantic knowledge base, and intelligent search and associated information recommendation functions can be realized.

Description

technical field [0001] The invention relates to the field of big data analysis of electric power, in particular to a method for constructing a knowledge map. Background technique [0002] The power grid infrastructure project has the characteristics of large scale, long cycle and many units involved. Therefore, it is determined that a large amount of data with rich sources and complex and diverse formats will be generated in the whole process of power grid engineering from design, construction to acceptance. Contains structured data, such as 3D geometric model logic model, etc.; semi-structured data such as equipment inventory, A list of materials, etc.; unstructured text data such as design instructions, experiment reports, etc. Information related to power grid topology, assets, and equipment can be obtained from the massive text data, and data in systems such as inspection, scheduling, and finance can be expanded and verified. [0003] However, since the establishment of...

AI Technical Summary

Problems solved by technology

[0003] However, since the establishment of these text data is mainly used for reading by engineering-related personnel, the format and semantics of most text content are not optimized, which is not conducive to computer processing and understanding

According to statistics, more than 70% of text information in infrastructure projects is stored in unstructured form, which greatly increases the difficulty and cost of applying these data to obtain information

Images