DDMF: A Method for Mining Relatively Important Nodes Based on Distance Distribution and Multi-Index Fusion

Abstract

In research on complex networks, mining relatively important nodes is a challenging and practical work. However, little research has been done on mining relatively important nodes in complex networks, and the existing relatively important node mining algorithms cannot take into account the indicators of both precision and applicability. Aiming at the scarcity of relatively important node mining algorithms and the limitations of existing algorithms, this paper proposes a relatively important node mining method based on distance distribution and multi-index fusion (DDMF). First, the distance distribution of each node is generated according to the shortest path between nodes in the network; then, the cosine similarity, Euclidean distance and relative entropy are fused, and the entropy weight method is used to calculate the weights of different indexes; Finally, by calculating the relative importance score of nodes in the network, the relatively important nodes are mined. Through verification and analysis on real network datasets in different fields, the results show that the DDMF method outperforms other relatively important node mining algorithms in precision, recall, and AUC value.

Ranking of Key Components of CNC Machine Tools Based on Complex Network

“…Wen [14] et al proposed a method based on the least square support vector machine (which allows the user to find the mapping rules between simple indicators and AHP evaluation) and showed the validity of the method in experiments on artificial and real networks. Zhao [15] et al proposed an important node mining method based on distance distribution and multiindex fusion (DDMF) to surpass the limitations of existing algorithms. In addition, the topology of the network impacts its stability.…”

Relative Node Importance in complex networks

“…Based on the random walk model, Wang et al [31] proposed the metric of RNI for finding Top-k most important nodes with respect to a set of query nodes in social networks, where the importance of node s with respect to node t is defined as the sum of path probabilities from s to t. Through measuring the relative importance of nodes in the criminal network, Alzaabi et al [32] proposed a forensic analysis system called crime investigation system, whereby the forensic investigators can determine the most influential members of a criminal group, or find the most related members to all known criminal members. Zhao et al [33] proposed a RNI mining method based on distance distribution and multi-index fusion, where the distance distribution is generated according to the shortest paths, and multiple indexes including the cosine similarity, Euclidean distance and relative entropy are fused to calculate the relative importance of nodes in the network. Although these metrics are proposed to measure the RNI, the systematical analysis, statistical characteristics, internal correlation principle and widespread applications of RNI are still the key problems to be researched.…”

“…Few concerns focus on the correlation among nodes, such as how much effects a node produces on other nodes in the network, who the most important person for another person is in social networks, and so on. Although some metrics of RNI [28][29][30][31][32][33] are propsed to meet some special requirements, the relavant theories and applications of RNI are still limited. Specifically, for example, the RNI can be applied to the following scenes:…”

Relative Node Importance in complex networks

“…Based on the random walk model, Wang et al [31] proposed the metric of RNI for finding Top-k most important nodes with respect to a set of query nodes in social networks, where the importance of node s with respect to node t is defined as the sum of path probabilities from s to t. Through measuring the relative importance of nodes in the criminal network, Alzaabi et al [32] proposed a forensic analysis system called crime investigation system, whereby the forensic investigators can determine the most influential members of a criminal group, or find the most related members to all known criminal members. Zhao et al [33] proposed a RNI mining method based on distance distribution and multi-index fusion, where the distance distribution is generated according to the shortest paths, and multiple indexes including the cosine similarity, Euclidean distance and relative entropy are fused to calculate the relative importance of nodes in the network. Although these metrics are proposed to measure the RNI, the systematical analysis, statistical characteristics, internal correlation principle and widespread applications of RNI are still the key problems to be researched.…”

“…Few concerns focus on the correlation among nodes, such as how much effects a node produces on other nodes in the network, who the most important person for another person is in social networks, and so on. Although some metrics of RNI [28][29][30][31][32][33] are propsed to meet some special requirements, the relavant theories and applications of RNI are still limited. Specifically, for example, the RNI can be applied to the following scenes:…”