Year: 2024
Type: conference-paper
Abstract: Computing the shortest path between vertices is a fundamental task in network analysis.This paper introduces an innovative strategy for selecting landmarks in a graph, which improves upon the traditional highest-degree selection method.Our approach captures the essential structural features of graphs, effectively reducing landmark overlap and clustering.Combined with partition querying and upper-bound pruning optimizations, our algorithm achieves a notable 20% -30% increase in query speed and is applicable to multi-million record datasets.This study offers new insights and solutions for enhancing query performance in large-scale graphstructured data, paving the way for further innovations in graph database optimization to meet the growing demands of large-scale graph data processing.
Authors: Zhipeng He, Dian Ouyang, Jianye Yang
Institution: Guangzhou University
Language: English
Topic: Graph Theory and Algorithms
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4411020775