基于大数据的新媒体网络平台转发预测算法研究  

Research on the Forwarding and Prediction Algorithm of the New Media Network Platform Based on Big Data

作　　者：吴蕾[1] 王海涛 鲁贻锦[1] WU Lei;WANG Haitao;LU Yijin(Anhui Police Officer Vocational College,Hefei 230031,China;School of Internet,Anhui University,Hefei 230039,China)

机构地区：[1]安徽警官职业学院,安徽合肥230031 [2]安徽大学互联网学院,安徽合肥230039

出　　处：《佳木斯大学学报(自然科学版)》2025年第2期46-49,共4页Journal of Jiamusi University:Natural Science Edition

基　　金：安徽省教育厅优秀青年教师培育一般项目(YQYB2023183)。

摘　　要：随着新媒体技术的更新换代,不少网络社交平台逐渐发展壮大,并随之产生了大量的文本数据。为了深度分析这些文本数据并提高转发预测的准确率,研究结合大数据技术与反向传播神经网络设计了一种新的舆情信息转发预测算法。首先结合大数据平台和改进关键字抽取算法处理舆情文本信息,接着利用自适应变异遗传算法优化后的反向传播神经网络设计了转发预测算法。研究结果表明,所设计的预测算法仅需迭代125次就能达到稳定状态,并且该算法在文娱类、科技类、财经类舆情信息中的预测准确率分别高达98.12%,97.44%,95.53%。由此可见,研究所设计的转发预测算法不仅具有较好的测试性能,同时在面对不同舆情信息时的预测准确率也较高,因此该算法可以为社交网络数据的深度分析提供技术支持。With the upgrading of new media technology,many network social media platforms have gradually developed and grown,and subsequently produced a large number of text data.In order to deeply analyze these text data and improve the accuracy of forwarding prediction,a new prediction algorithm of public opinion information forwarding is designed by combining big data technology and backpropagation neural network.First of all,we combined the big data platform and improved the key word extraction algorithm,and then the forwarding prediction algorithm is designed by using the back propagation neural network optimized by the adaptive variant genetic algorithm.The results show that the designed prediction algorithm can reach a stable state after only 125 iterations,and the prediction accuracy of the algorithm in the public opinion information of entertainment,science and technology and finance is as high as 98.12%,97.44%and 95.53%respectively.It can be seen that the forwarding prediction algorithm designed by the institute not only has good test performance,but also has high prediction accuracy in the face of different public opinion information,so the algorithm can provide technical support for the in-depth analysis of social network data.

关 键 词：BP 社交网络 新媒体 舆情信息 预测算法 文本挖掘技术 

分 类 号：R283[医药卫生—中药学]