基于熵权TOPSIS的社会网络影响力最大化研究

Research on social network influence maximization based on entropy TOPSIS

在影响力最大化问题中,针对启发式方法精度不足和贪婪方法时间过载的问题,提出一种基于多属性决策方法的影响力最大化算法。首先,从社会网络节点的影响传播、节点之间的影响重叠和节点的信任度角度选取节点的重要性评价指标。然后,建立基于熵权TOPSIS的社会网络节点重要性评价模型,通过模型选择影响范围最广、与当前种子集的重叠最小且信任度最高的节点。最后,构建算法,并通过实验验证算法的性能。实验结果表明,与传统影响力最大化算法相比,所提算法在传播范围与时间效率上取得了较好的折中。

In the influence maximization problem,aiming at the problems of insufficient accuracy of heuristic methods and time overload of greedy methods,this study proposed an influence maximization algorithm based on multi-attribute decision-making method.Firstly,it selected the importance evaluation indexes of nodes from the perspective of the influence spread of nodes,the influence overlap between nodes and the trust degree of nodes.Then,it established a social network node importance evaluation model based on entropy weight TOPSIS,and selected the node that had the widest influence range,the smallest overlap with the current seed set and the highest trust degree through the model.Finally,this study constructed an algorithm and verified the performance of the algorithm through experiments.The experimental results show that compared with the traditional influence maximization algorithm,the proposed algorithm achieves a better compromise between the spread range and time efficiency.