Abstract extraction method and system for homogeneous relationship large graph

Abstract

The invention provides an abstract extraction method and system for a homogeneous relation large graph, and the method comprises the steps of obtaining relation graph data to be subjected to abstract extraction as current graph data, enabling the relation graph data to be the homogeneous relation large graph, and enabling each node in the current graph data to serve as a super point; according to the adjacent matrix of the current graph data, grouping nodes in the current graph data through locality sensitive hashing; randomly selecting a plurality of super-point pairs from the group, respectively calculating the difference between the combined super-point pairs and the relational graph data, and selecting the super-point pair with the minimum difference for combination to obtain reconstructed graph data; and outputting the reconstructed image data as an abstract extraction result.

Description

technical field [0001] The invention relates to the field of data mining, in particular to a rapid summary and reconstruction technology and device for large homogeneous relationship graphs. Background technique [0002] At present, social media has surpassed search engines to become the largest source of Internet traffic, accounting for 46% and 40% respectively. Relational graph data has become a common data used in many sciences and engineering. The graph can be expressed as such a structure, that is, the graph G = (V, E) is a pair of sets: a set of nodes V represents entities and a set of edges E Represents relationships or connections between entities. In computer science, networks consist of nodes and edges; in social science, the corresponding terms are actors and relationships, which are used synonymously in this paper. As of the first quarter of 2020, the combined monthly active accounts of WeChat and WeChat reached 1.2025 billion, which means that WeChat has offic...

AI Technical Summary

Problems solved by technology

[0006] The methods mentioned above are mainly focused on static simple graphs and applied to a certain type of graph data, and cannot be universally applicable

At the same time, the above method needs to calculate the relationship between each pair of nodes in order to summarize the graph data. Although there are some methods to optimize and speed up the above calculation process, the computational complexity is still high, especially in the face of large graphs. When collecting data, these methods generally have shortcomings such as low efficiency, time-consuming, and need to occupy a lot of memory.

Images