What is it about?
Community detection, or clustering, is the problem of identifying natural divisions in networks/graphs. Efficient parallel algorithms for identifying such divisions is critical in a number of applications, where the size of datasets have reached significant scales. This paper presents one of the most efficient implementations of the Leiden algorithm, a high quality community detection method.
Featured Image
Photo by Nareeta Martin on Unsplash
Why is it important?
Communities, also known as clusters, shed light on the organization and functionality of the network. This problem with applications in topic discovery, protein annotation, recommendation systems, and targeted advertising. One of the difficulties in the community detection problem is the lack of apriori knowledge on the number and size distribution of communities. The Louvain method is a popular heuristic-based approach for community detection. Despite its popularity, the Louvain method has been observed to produce internally-disconnected and badly connected communities. To address these shortcomings, the Leiden algorithm has been proposed.
Perspectives
In this paper, we show that, for irregular algorithms, efficient multicore algorithms can outperform GPUs.
Dr. Subhajit Sahu
SRM University
Read the Original
This page is a summary of: Fast Leiden Algorithm for Community Detection in Shared Memory Setting, August 2024, ACM (Association for Computing Machinery),
DOI: 10.1145/3673038.3673146.
You can read the full text:
Contributors
The following have contributed to this page
