Frequent sub-graph mining method aiming at individual uncertain graph

Abstract

The invention discloses a frequent sub-graph mining method aiming at an individual uncertain graph. The method comprises the steps of: acquiring the individual uncertain graph; enumerating all sub-graphs of the individual uncertain graph according to the individual uncertain graph; specifying partial implication graphs in all implication graphs of the individual uncertain graph as the graph samples; respectively calculating the existence probability of each sample graph of the individual uncertain graph as well as the support degree of each sub-graph on the graph sample of the individual uncertain graph; judging whether the sub-graph is the frequent sub-graph according to the expected support degree of each sub-graph on the graph sample of the individual uncertain graph and the existence probability of each sample graph of the individual uncertain graph; outputting all frequent sub-graphs. According to the frequent sub-graph mining method aiming at the individual uncertain graph, a frequent sub-graph mining technology can be used on the single uncertain graph by dividing the single uncertain graph into a plurality of implication graphs and regarding the implication graphs as certain graphs to sample and calculate the expected support degree, so that the technical gap in the field is filled.

Description

technical field [0001] The invention relates to graph mining technology, in particular, to a frequent subgraph mining method for a single uncertain graph. Background technique [0002] Uncertainty is an inherent property in real-world applications, whether for internal or external sources. For example, in a cooperative social network, we may not be able to clearly assert that Bill and Matthew have a good cooperative relationship using the currently available information, and we usually use probability to measure the possibility of such a cooperative relationship. Assuming that the probability of such a relationship exists is p, the value of p is manually determined by experts in the field from the available information, or is automatically generated by information extraction or generation rules. In today's era of big data, there is a stronger demand for managing uncertain data, so there are various data of different quality. In particular, we focus on uncertain graphs, esp...

AI Technical Summary

Problems solved by technology

[0006] In view of the lack of frequent subgraph mining technical solutions for a single uncertain graph in the prior art, the purpose of the present invention is to propose a frequent subgraph mining method for a single uncertain graph, which can allow frequent subgraph mining for a single uncertain graph. Subgraph mining fills the technical gap in this field

Images