Original Label Propagation Algorithm Research Articles

LPA-MNI: An Improved Label Propagation Algorithm Based on Modularity and Node Importance for Community Detection.

Community detection is of great significance in understanding the structure of the network. Label propagation algorithm (LPA) is a classical and effective method, but it has the problems of randomness and instability. An improved label propagation algorithm named LPA-MNI is proposed in this study by combining the modularity function and node importance with the original LPA. LPA-MNI first identify the initial communities according to the value of modularity. Subsequently, the label propagation is used to cluster the remaining nodes that have not been assigned to initial communities. Meanwhile, node importance is used to improve the node order of label updating and the mechanism of label selecting when multiple labels are contained by the maximum number of nodes. Extensive experiments are performed on twelve real-world networks and eight groups of synthetic networks, and the results show that LPA-MNI has better accuracy, higher modularity, and more reasonable community numbers when compared with other six algorithms. In addition, LPA-MNI is shown to be more robust than the traditional LPA algorithm.

Graph-based label propagation algorithm for community detection

Community detection is one of the most important topics in complex network analysis. Among a variety of approaches for detecting communities, the label propagation algorithm (LPA) is the simplest and time-efficient approach. However, the original label propagation algorithm is not stable due to the randomness in its propagation process. In this paper, we propose a graph-based label propagation algorithm (GLPA) to detect communities incorporating the node similarity and connectivity information during the propagation of the labels. First, we define node similarity between adjacent nodes, and change each node’s label to that of its most similar neighbor node. Based on the label propagation process, GLPA constructs a label propagation graph to get candidate communities. Then, GLPA calculates the connected components of the label propagation graph. Each connected component is treated as a candidate community in the next step. Second, GLPA constructs a weighted graph to obtain final communities, in which each connected component are treated as a super-node, and the number of edges lying between the corresponding components as the weight of edges. We compute the merging factor for each node in the weighted graph and merge super nodes with higher merging factor to its most similar node iteratively to reach the maximum complementary entropy. Compared with 8 other classical community detection algorithms on LFR artificial networks and 12 real world networks, the proposed algorithm GLPA shows preferable performance on stability, NMI, ARI, modularity.

A Link-Based Similarity for Improving Community Detection Based on Label Propagation Algorithm

Community structure is one of the most best-known properties of complex networks. Finding communities help us analyze networks from a mesoscopic viewpoints instead of microscopic or macroscopic one. It helps to understand behavior grouping. Various community detection algorithms have been proposed with some shortcomings in time and space complexity, accuracy, or stability. Label Propagation Algorithm (LPA) is a popular method used for finding communities in an almost-linear time-consuming process. However, its performance is not satisfactory in some metrics such as accuracy and stability. In this paper, a new modified version of LPA is proposed to improve the stability and accuracy of the LPA by defining two concepts -nodes and link strength based on semi-local similarity-, while preserving its simplicity. In the proposed method a new initial node selection strategy, namely the tiebreak strategy, updating order and rule update are presented to solve the random behavior problem of original LPA. The proposed algorithm is evaluated on artificial and real networks. The experiments show that the proposed algorithm is close to linear time complexity with better accuracy than the original LPA and other compared methods. Furthermore, the proposed algorithm has the robustness and stability advantages while the original LPA does not have these features.

LP-LPA: A link influence-based label propagation algorithm for discovering community structures in networks

Community detection is an essential approach for analyzing the structural and functional properties of complex networks. Although many community detection algorithms have been recently presented, most of them are weak and limited in different ways. Label Propagation Algorithm (LPA) is a well-known and efficient community detection technique which is characterized by the merits of nearly-linear running time and easy implementation. However, LPA has some significant problems such as instability, randomness, and monster community detection. In this paper, an algorithm, namely node’s label influence policy for label propagation algorithm (LP-LPA) was proposed for detecting efficient community structures. LP-LPA measures link strength value for edges and nodes’ label influence value for nodes in a new label propagation strategy with preference on link strength and for initial nodes selection, avoid of random behavior in tiebreak states, and efficient updating order and rule update. These procedures can sort out the randomness issue in an original LPA and stabilize the discovered communities in all runs of the same network. Experiments on synthetic networks and a wide range of real-world social networks indicated that the proposed method achieves significant accuracy and high stability. Indeed, it can obviously solve monster community problem with regard to detecting communities in networks.

LPA-CBD an improved label propagation algorithm based on community belonging degree for community detection

In order to deal with stochasticity in center node selection and instability in community detection of label propagation algorithm, this paper proposes an improved label propagation algorithm named label propagation algorithm based on community belonging degree (LPA-CBD) that employs community belonging degree to determine the number and the center of community. The general process of LPA-CBD is that the initial community is identified by the nodes with the maximum degree, and then it is optimized or expanded by community belonging degree. After getting the rough structure of network community, the remaining nodes are labeled by using label propagation algorithm. The experimental results on 10 real-world networks and three synthetic networks show that LPA-CBD achieves reasonable community number, better algorithm accuracy and higher modularity compared with other four prominent algorithms. Moreover, the proposed algorithm not only has lower algorithm complexity and higher community detection quality, but also improves the stability of the original label propagation algorithm.

CenLP: A centrality-based label propagation algorithm for community detection in networks

Community detection is an important work for discovering the structure and features of complex networks. Many existing methods are sensitive to critical user-dependent parameters or time-consuming in practice. In this paper, we propose a novel label propagation algorithm, called CenLP (Centrality-based Label Propagation). The algorithm introduces a new function to measure the centrality of nodes quantitatively without any user interaction by calculating the local density and the similarity with higher density neighbors for each node. Based on the centrality of nodes, we present a new label propagation algorithm with specific update order and node preference to uncover communities in large-scale networks automatically without imposing any prior restriction. Experiments on both real-world and synthetic networks manifest our algorithm retains the simplicity, effectiveness, and scalability of the original label propagation algorithm and becomes more robust and accurate. Extensive experiments demonstrate the superior performance of our algorithm over the baseline methods. Moreover, our detailed experimental evaluation on real-world networks indicates that our algorithm can effectively measure the centrality of nodes in social networks.

A Fast and Accurate Algorithm for Detecting Community Structure in Social Networks

Studies of community structure and evolution in large social networks require fast and accurate algorithms for community detection. Among the existing algorithms for community detection, the label propagation algorithm (LPA) and the Newman modularity Q algorithm (NMA) have been widely used and studied in the community detection in large social networks, since the LPA has the advantages of near-linear running time, easy implementation and without requiring parameters, and the NMA is a relatively fast algorithm and has a clear metrics to measure community structure. However, the LPA has the shortcomings that the result of the community detection is instable and has a low quality. At the same time, disadvantages of the NMA are that it bases its decisions on purely local information about individual communities and gets the local optimal solution. In this paper, combined with these two algorithms, we propose a new community detection algorithm (LP-NMA), which extends the above two algorithms (the LPA and the NMA is a special case of the new algorithm respectively). The new algorithm not only retains the advantages of these two algorithms, but also has improved the stability and quality of community detection. Experiments on real social networks have proved that this method is better than the original LPA and NMA.

Using Prior Knowledge for Community Detection by Label Propagation Algorithm

Community detection is an important approach to analyze and understand the organization or unit structure of the complex networks. By comparing the existing community detection algorithms, the label propagation algorithm (LPA) shows prominent operation speed and qualifies near linear time complexity. However, original LPA algorithm only uses the topological structure to guide the community detection process, failing to improve the quality of community detection when extra information offered. In this paper, we combine the prior information with topological structure to guide the community detection process. During the label propagation process, we proposed a new label update principle, making a node absorb its neighbor label information depending on the label distribution. The experimental results both on real networks and artificial networks show that the improved algorithm not only inherits the characteristic of rapid speed, but also improves the quality of community detection. Moreover, the improved algorithm still has the feature of near linear time complexity.