基于固定邻域规模的动态网络影响力最大化探测算法

Dynamic network influence maximization detection algorithm based on regular area scale

以往对影响力最大化问题的研究大多是基于静态图进行优化研究,但在现实中,网络数据量随着时间不断增加,系统不可能实时获取到整个网络中节点之间的连接情况。在传统MaxG探测模型的基础上,采用固定邻域规模和节点邻域层级相结合的方式计算节点影响力大小,提出了新的动态网络探测算法RAS-MaxG(regular area scale-MaxG),解决了传统探测算法由于采用度来衡量节点影响力值所导致的节点之间区分性差的问题。最后通过在真实数据集上的实验对比,验证了所提算法在最终影响力覆盖范围方面具有更好的性能表现。

Previous research on influence maximization are mostly based on static graph optimization,but in reality,the network data volume increase rapidly with time,so the system cannot obtain the connection between nodes in the whole network in real time.Based on the traditional MaxG detection model,this paper proposed the RAS-MaxG detection algorithm,which combined the regular area scale and node neighborhood level to calculate the node influence.Finally,the experimental results on real datasets show that the proposed detection algorithm has better performance in the final influence coverage.