基于智能算法的食品安全网络舆情监测方法研究

Research on Public Opinion Monitoring Method of Food Safety Network Based on Intelligent Algorithm

近年来,我国频繁出现了各类食品安全话题,特别是在互联网环境下,具有传播速度快、影响深远等特点。当前已经进入一个大数据时代,对食品安全网络舆情的监测任务迫在眉睫。利用Single-Pass话题检测算法可以开展话题检测及文本分析,但是在时效性和准确性上还存在一定的不足,因而此次提出了引入Storm分布式框架及自编码神经网络来对算法进行优化改进。通过对优化算法的性能测试可以看出,经过改进后Single-Passs算法在时效性和聚类精度上都有了显著的提升。基于优化改进算法的食品安全网络监测系统进行网络话题跟踪、检测等,能够较好实现热点话题的跟踪,并绘制出了热点话题的趋势图。

In recent years,all kinds of food safety topics have appeared frequently in China,especially in the Internet environment,which has the characteristics of fast transmission speed and far-reaching influence.At present has entered a big data era,the monitoring task of food safety network public opinion is urgent.Using Single-Pass topic detection algorithm can carry out topic detection and text analysis,but there are still some shortcomings in timeliness and accuracy,so this paper proposes to introduce Storm distributed framework and self-coding neural network to optimize and improve the algorithm.By testing the performance of the optimization algorithm,it can be seen that the improved Sin gle-Passs algorithm has improved significantly in timeliness and clustering accuracy.The food safety network monitoring system based on the optimized and improved algorithm can track and detect the hot topics,and draw the trend map of the hot topics.