基于并行阻塞策略的社交网络舆情传播控制方法

Blocking Strategy Based Social Network Public Opinion Communication Control Method

近年来,随着社交网络的不断普及,负面舆情信息对经济、社会和文化等造成的威胁不断放大.现有方法多通过剔除节点或边集实现舆情传播控制,却难以适应现今规模巨大的社交网络环境.基于此,基于阻塞自回避行走算法提出了融合边与节点集移除的并行舆情传播控制方法,并基于GPU框架应用成本约束函数对算法进行改进,提高了所提算法在大规模社交网络图中的适用性,且使用改进前推回代算法进行求解.最后,基于多种真实社交网络数据集进行实验,结果表明:所提CN-OPC法能对社交网络舆情传播进行有效控制,控制效率与精度较经典算法均有较大提升,这说明并行阻塞策略可用于大规模社交网络舆情传播控制,未来可进一步尝试多策略并行的控制方法在社交网络环境中的应用.

In recent years,with the continuous popularization of social networks,the threat of negative public opinion information to economy,society and culture is constantly amplified.The existing methods mostly eliminate nodes or edge sets to achieve public opinion communication control,but it is difficult to adapt to today's huge social network environment.Based on this,a parallel public opinion propagation control method combining edge and node set removal is proposed based on the blocking self-avoidance walking algorithm,and the algorithm is improved by using the cost constraint function based on the GPU framework to make it suitable for large-scale social network graph,and the improved push-back algorithm is used to solve the problem.Finally,experiments are conducted based on a variety of real social network datasets,and the results show that:The proposed CN-OPC method can effectively control the spread of public opinion in social networks,and the control efficiency and accuracy are greatly improved compared with the classical algorithm.This indicates that the parallel blocking strategy can be used to control the spread of public opinion in large-scale social networks,and the application of the multi-strategy parallel control method in the social network environment can be further attempted in the future.