基于支持向量机的网络舆情危机预警探究  

Study of support vector machine based on network public opinion crisis warning

作　　者：孙亮[1] Sun Liang

机构地区：[1]兰州文理学院数字媒体学院,甘肃兰州730000

出　　处：《自动化与仪器仪表》2016年第11期138-139,共2页Automation & Instrumentation

摘　　要：针对信息化条件下的网络舆情分析需要,结合计算机理论,提出一种基于支持向量机算法的网络舆情预警模型。利用层次聚类算法对爬行器搜集到的数据进行预处理,从而提高数据的质量;结合网络舆情的相关预警指标,利用灰色理论模型构建网络舆情预测模型,并通过SVM支持向量机对预测的结果进行修正。最后通过仿真测试,与传统的预测模型相比,该算法具有较高的精确度。in view of the network public opinion under the condition of information analysis,combining the theory of computer,proposes a network public opinion warning model based on support vector machine algorithm.Using hierarchical clustering algorithm of crawler to collect the data pretreatment,so as to improve the data quality;combined with the network of public opinion related to early warning indicators,construction of network public opinion prediction model based on grey theory model,and the SVM support vector machine to predict the results of modified.Finally,compared with the traditional prediction model,the algorithm has higher accuracy.

关 键 词：支持向量机 舆情危机 预警 聚类算法 网络爬行器 

分 类 号：TP393.092[自动化与计算机技术—计算机应用技术]