In recent years, with the strong development of information technology, detecting communities in large real networks is a very important issue which is of interest to many scientists. Community detection in large real networks with millions of nodes is often difficult. To solve this problem, many online community search algorithms have been proposed with many different approaches. One of the approaches is to coordinate the vertices of the graph and build a reasonable distance between those vertices. It has been observed that vertices in the same community have approximately the same probability of reaching other vertices through a random walk. Based on this principle, the authors propose a way to coordinate vertices and build distances between vertices in the graph that reduces computational complexity compared to existing techniques. This approach involves representing peaks as vectors and using the K-means++ algorithm for community detection, whose effectiveness is evaluated through...
In recent years, with the strong development of information technology, detecting communities in large real networks is a very important issue which is of interest to many scientists. Community detection in large real networks with millions of nodes is often difficult. To solve this problem, many online community search algorithms have been proposed with many different approaches. One of the approaches is to coordinate the vertices of the graph and build a reasonable distance between those vertices. It has been observed that vertices in the same community have approximately the same probability of reaching other vertices through a random walk. Based on this principle, the authors propose a way to coordinate vertices and build distances between vertices in the graph that reduces computational complexity compared to existing techniques. This approach involves representing peaks as vectors and using the K-means++ algorithm for community detection, whose effectiveness is evaluated through experimental results presented.