| How to study the complex networks control,has become an important task.Also,it has an important practical value in practice,especially in all kinds of large or complex networks.In order to reduce the cost control,the network can be targeted to analyze its properties,or we can develop the right strategy and measures to guide the network system to develop in the direction of what we want.Compared to the other nodes,the important nodes in the network affect the network structure and function in a larger extent.These nodes also can spread more information in the network.So this work needs to think about the following two questions in the target control:(I)How to control the selected node that can achieve control goal?(II)How to prevent the spread of certain information in the network by blocking only the few nodes?In order to pick out the important node for the target control,and study its importance of information propagation in the network,this paper made the following two aspects:(1)Target control,it aims to study how to choose and control the part of nodes in the network,and it has been mainly adopted random and local selection without considering the importance of nodes.Since the node importance evaluation index of existing complex network is single,the paper adopts a node importance evaluation method based on multiple-attribute decision from different angels by using several node importance indices.In order to select important nodes for the target control,we give different weights to evaluate and rank the nodes.The experimental results show that the method is able to select less driver nodes in both artificial data and real data sets.(2)Based on the theory of target control and greedy algorithm in complex networks,this article studies the target control from the perspective of propagation immunization of the complex network.The methord first selects important nodes according to the comprehensive index to insulates them,and then random selects the target nodes in the entire network for the target control.Using this method can isolate part of the transmission of information in the network,and then we analyze the results generated after applying immunization strategy in the both of real network and artificial network in the target control.As a result,it is concluded that comprehensive indicator removes edge information of nodes more than the single index in the network,so implementing the immunization strategy by the comprehensive index can get the more driver nodes.In a word,the research of identifying significant nodes from two aspects are studied in this paper.Experimental results show the effectiveness of the algorithms on both artificial data and real data sets.The research will provide new ideas for the identification of important nodes and propagation immunization in the target control,it has application values in some domains such as knowledge discovery and network control. |
