Large-scale graphic constraint linkage path searching algorithm in high-dimensional vector space

Abstract

The invention relates to a large-scale graphic constraint linkage path searching algorithm in a high-dimensional vector space. The algorithm mainly includes the following steps of projection screening; dimensionality reduction deletion and selection; node deletion and selection; node subset high-dimensional spatial integration, wherein constraint linkage paths are output. By means of the algorithm, a node similarity linkage algorithm based on distance in the vector space is improved, a gridding technique and constraint linkage characters are introduced, a great quantity of invalid nodes are screened out, and the calculation complexity is reduced; a Map-Reduce framework is introduced into the algorithm, the path of each node is calculated by means of a four-stage deletion and selection strategy, an optional node set meeting the constraint conditions is found, a result set is called circularly to return to a reachable constraint path, unnecessary text copying costs and unnecessary path calculation costs are reduced, and the algorithm has high efficiency and a low error rate. Meanwhile, certain improvement and innovation are conducted on the algorithm during processing of the high-dimensional data space, unnecessary calculation processing is decreased through various dimension reduction means, the processing execution capability of the algorithm is improved, and therefore the algorithm can give answers and make responses to the questions of users more quickly.

Description

technical field [0001] The invention belongs to the field of big data, and in particular relates to the design of a processing method for a large-scale graph data set in a vector space, in particular to the design of a query algorithm for a large-scale graph constraint connection path in a high-dimensional vector space. Background technique [0002] In recent years, with the continuous development of new services such as big data and cloud computing, applications related to spatial location have gradually increased, and the scale of spatial map data is growing and accumulating at an unprecedented rate. How to find satisfying users in high-dimensional spatial map datasets? The optimal path result set for demand is a research hotspot in the field of large-scale graphs. The Map-Reduce framework provides effective means for batch processing of large-scale data. In the Map stage, according to the self-similarity of the high-dimensional vector space data set, the size of the rela...

AI Technical Summary

Problems solved by technology

[0004] In order to solve the problem that the analysis operation of a large-scale graph data set in a high-dimensional space requires high calculation costs, the present invention provides an algorithm with low computational complexity and high execution capability in a large-scale graph constrained connection path in a high-dimensional vector space query algorithm

Images