基于人工智能的网络舆情大数据传播特征挖掘系统  

AI-based propagation feature mining system for network public opinion big data

作　　者：罗平 武斌 LUO Ping;WU Bin(Shanghai University of Finance and Economics Zhejiang College,Jinhua 321013,China)

机构地区：[1]上海财经大学浙江学院

出　　处：《现代电子技术》2020年第4期176-179,共4页Modern Electronics Technique

基　　金：浙江省教育厅一般基金(Y201636713)

摘　　要：为了有效解决舆情数据传播中特征挖掘技术面临的特征挖掘滞后问题,设计基于人工智能的网络舆情大数据传播特征挖掘系统。设计网络信息传感提取模块,利用O/I顺向电流指向性电路节点端口和电机控制性匹配结构构建外围控制电路以及带状数据分布式交互接口。设计舆论数据人工智能入网协议来转换当前的网络数据格式,并对其进行舆情二次语义处理。文中使用人工智能技术,依靠Hadoop平台内部学习库中的Ma_Orduse算法和K-means算法通过有效数据文本聚类获取特征键值,以中枢结构的形式传递回中央控制器,实现当前网络舆情大数据传播特征挖掘。实验数据表明应用该挖掘系统后,舆情数据特征读入延迟降低32%,读出延迟降低27%,可以有效缓解特征挖掘滞后问题。An AI⁃based propagation feature mining system for network public opinion big data is designed to effectively solve the lag problem of feature mining faced by the feature mining technology of public opinion data propagation.The network information sensing and extraction module is designed,and the peripheral control circuit and banded data distributed interactive interface are constructed with O/I forward current directivity circuit node port and motor control matching structure.The artificial intelligence network protocol of public opinion data is designed to convert the current network data format and carry out the secondary semantic processing of public opinion of it.The characteristic key values are obtained by effective data text clustering by means of the Ma_Orduse algorithm and K⁃means algorithm in the internal learning library of Hadoop platform with the artificial intelligence technology,which are transferred back to the central controller in the form of central structure to realize the feature mining of current network big data transmission of public opinion.The experimental data show that the application of the mining system,the read⁃in delay of public opinion data features is reduced by 32%,and the read⁃out delay is reduced by 27%,which can effectively alleviate the lag problem of feature mining.

关 键 词：舆情大数据 特征挖掘 人工智能 系统设计 数据分析 语义处理 

分 类 号：TN911-34[电子电信—通信与信息系统] F272[电子电信—信息与通信工程]