Experimental Evaluation of Distributed k-Core Decomposition

TL;DR

This paper presents an experimental evaluation of a distributed algorithm for k-core decomposition in large graphs, using simulation to analyze its performance in terms of runtime and message passing.

Contribution

It introduces a simulation-based experimental framework for assessing distributed k-core decomposition algorithms on large-scale graphs.

Findings

Simulation effectively evaluates runtime and message passing.

Distributed approach overcomes memory constraints of sequential algorithms.

Golang's Goroutines facilitate scalable simulation.

Abstract

Given an undirected graph, the $k$-core is a subgraph in which each node has at least $k$ connections. This is widely used in graph analytics to identify core subgraphs within a larger graph. The sequential $k$-core decomposition algorithm faces limitations due to memory constraints, and many data graphs are inherently distributed. A distributed approach is proposed to overcome limitations by allowing each vertex to compute its core number independently using only local information. This work explores the experimental evaluation of a distributed $k$-core decomposition algorithm. By assuming that each vertex is a client as a single computing unit, we simulate the process using Golang, leveraging its Goroutines and message passing. Since real-world data graphs can be large with millions of vertices, it is expensive to build a distributed environment with millions of clients if experiments were run in a real distributed environment. Therefore, our experimental simulation can effectively evaluate the running time and message passing for the distributed $k$-core decomposition.