Large-scale operation and maintenance knowledge graph storage and multi-granularity visualization method

Abstract

The invention discloses a large-scale operation and maintenance knowledge graph storage and multi-granularity visualization method, which comprises the following steps of: 1) using an X-Engine engine as a historical data storage core to complete knowledge graph data storage of massive entities, attributes and relationships; 2) clustering the knowledge graph data according to an LPA algorithm, and fusing entities with relatively high similarity into a group or a community; and 3) according to the step 1), the step 2) and the query condition of the user, performing query processing and data request operation on the intelligent operation and maintenance knowledge graph at the server side, performing multi-granularity graphic rendering at the terminal according to a result, and displaying the result. An X-Engine database storage engine used in the method is higher in writing performance, the impact of Compaction actions in a traditional LSM-tree framework on system resources is reduced, the system performance is kept stable, knowledge graph data is clustered through an LPA algorithm, through multi-granularity visualization, the data size displayed in a browser is greatly reduced, the disorder phenomenon is avoided, and the reliability of the system is improved. And the rendering performance is improved.

Description

technical field [0001] The invention relates to the field of storage and visualization of knowledge graphs, in particular to a large-scale operation and maintenance knowledge graph storage and multi-granularity visualization method. Background technique [0002] Label Propagation Algorithm (LPA algorithm) is a community partition algorithm based on label propagation and clustering. The basic idea is: first initialize a unique unique initial label for each node in the network; then in each iteration, the node will choose the label that appears most frequently in its neighbors, and use this label to change the label of the node itself, if the neighbor If the same label appears multiple times in the same label, randomly select one of the same labels to replace the label of the node itself; then iterate continuously until the labels of all nodes no longer change, and cluster the nodes holding the same label as a local community. Its advantages are simple logic, low time comple...

AI Technical Summary

Problems solved by technology

Second, the network topology structure of the operation and maintenance object is more complex, especially the dynamic change of the topology relationship of the information system operation is greater

Specifically, the main difficulties include: First, it is difficult to render a large number of nodes and edges. Visualizing the entire knowledge graph or a large amount of information on a limited-space display is likely to cause visual confusion, and users cannot quickly focus on useful information, and the drawing efficiency of large-scale graphs low, which affects the fluency of human-computer interaction; second, the front-end computing performance pressure brought by rich clients is difficult, and the amount of data in large-scale visual rendering is huge. The content of the browser is overflowing; the third is the difficulty of interactive processing such as data filtering, location picking and combination. In a pile of massive data maps, too much information will easily confuse users, and less information will lead to no useful analysis results

Images