A Simple Acceleration Method for the Louvain Algorithm

  title={A Simple Acceleration Method for the Louvain Algorithm},
  author={Naoto Ozaki and Hiroshi Tezuka and Mary Inaba},
  journal={International Journal of Computer and Electrical Engineering},
The Louvain algorithm is well known for its high speed for detecting community structure in networks. In this paper, first, we analyze the Louvain algorithm as the preliminary experiment to uncover the processes that cause wasted computational time and their characteristics. Then based on this, we propose the Louvain Prune algorithm. The experiments show that the Louvain Prune algorithm significantly reduces computational time by up to 90%, and retains almost the same quality as the original… 

Accelerating Louvain community detection algorithm on graphic processing unit

An adaptive CUDA Louvain method (ACLM) algorithm that benefits from the graphic processing unit (GPU) and uses the shared memory in GPU, as well as the optimal number of threads in the GPU blocks to minimize parallelization overhead and accelerate the calculation of modularity parameters.

From Louvain to Leiden: guaranteeing well-connected communities

The Leiden algorithm is found to be faster than the Louvain algorithm and uncovers better partitions, in addition to providing explicit guarantees on communities that are guaranteed to be connected.

Improving Louvain Algorithm by Leveraging Cliques for Community Detection

  • Elaf Adel AbbasH. N. Nawaf
  • Computer Science
    2020 International Conference on Computer Science and Software Engineering (CSASE)
  • 2020
The traditional Louvain algorithm is used for community detection as a suitable algorithm, since it provides fast, efficient and robust community detection on large static networks, but the high computing complexity of this algorithm is a motivation of this work.

Agglomerative likelihood clustering

It is argued that ALC can reduce computation time costs and resource usage costs for large scale clustering for time-series applications while being serialized, and hence has no obvious parallelization requirement.

Prune the Unnecessary: Parallel Pull-Push Louvain Algorithms with Automatic Edge Pruning

A collection of parallel Louvain algorithms that prune many edges and vertices, speeding up convergence by an order of magnitude over the previously best-known implementation of theLouvain algorithm from Grappolo, while producing similar or better results.

Graph-Based Semantic Segmentation

This work assesses the Louvain algorithm with the COLD-Freiburg dataset to create the semantic map and compute the communities and demonstrates that lighting conditions do not affect the system’s ability to categorize places.

Community detection using fast low-cardinality semidefinite programming

This paper proposes a new class of low-cardinality algorithm that generalizes the local update to maximize a semidefinite relaxation derived from max-k-cut, which is scalable, empirically achieves the global semideFinite optimality for small cases, and outperforms the state-of-the-art algorithms in real-world datasets with little additional time cost.

Machine Learning Approach to Community Detection in a High-Entropy Alloy Interaction Network

This paper attempts to build interaction networks based on high-entropy alloy (HEA) descriptors in order to discover HEA communities with similar functionality and shows that the method with the PSO-based community detection algorithm can achieve alloy communities with an average accuracy improvement of 0.26 compared to the Louvain algorithm.



Improving the Louvain Algorithm for Community Detection with Modularity Maximization

This paper presents an enhancement of the well-known Louvain algorithm for community detection with modularity maximization by including a uncoarsening phase, leading to a full multi-level method.

Faster unfolding of communities: speeding up the Louvain algorithm

  • V. Traag
  • Computer Science
    Physical review. E, Statistical, nonlinear, and soft matter physics
  • 2015
This work suggests to consider moving nodes to a random neighbor community, instead of the best neighborcommunity, which reduces the theoretical runtime complexity from O(m) to O(nlog〈k〉) in networks with a clear community structure.

A smart local moving algorithm for large-scale modularity-based community detection

It is shown that the proposed smart local moving algorithm identifies community structures with higher modularity values than other algorithms for large-scale modularity optimization, among which the popular “Louvain algorithm”.

Fast unfolding of communities in large networks

This work proposes a heuristic method that is shown to outperform all other known community detection methods in terms of computation time and the quality of the communities detected is very good, as measured by the so-called modularity.

Accelerating Community Detection by Using K-core Subgraphs

This work proposes a framework that accelerates community detection by applying an expensive algorithm (modularity optimization, the Louvain method, spectral clustering, etc.) to the K-core and then using an inexpensive heuristic to infer community labels for the remaining nodes.

Fast parallel algorithm for unfolding of communities in large graphs

The proposed distributed memory parallel algorithm targets the costly first iteration of the initial method by parallelizing it and achieves up to ≈5× performance improvement as compared to the sequential version while not compromising the correctness of the final result.

An ensemble learning strategy for graph clustering

It is shown, that the quality of the initial weak clusterings is of minor importance for the final result of the scheme if the authors iterate the process of restarting from maximal overlaps, and its search behavior is linked to global analysis.

Benchmark graphs for testing community detection algorithms.

This work introduces a class of benchmark graphs, that account for the heterogeneity in the distributions of node degrees and of community sizes, and uses this benchmark to test two popular methods of community detection, modularity optimization, and Potts model clustering.

KONECT: the Koblenz network collection

KONECT's taxonomy of networks datasets is described, an overview of the datasets included, a review of the supported statistics and plots, and the project's role in the area of web science and network science are discussed.

Resolution limit in community detection

It is found that modularity optimization may fail to identify modules smaller than a scale which depends on the total size of the network and on the degree of interconnectedness of the modules, even in cases where modules are unambiguously defined.