利用BP神经网络对网络舆情进行预警

BP Neural Network For Early Warning Network Public Opinion

随着近年来中国网民数量众多而网络给予了众多网民言论自由,因此网络舆情越来越被关注,对网络舆情的预警也越来越迫切。该文采用了五个指标对“厦门PX事件”、“昆明PX事件”和“宁波PX事件”进行量化描述,通过Matlab建立BP神经网络识别模型中进行学习,然后再对“宁波PX事件”进行预警识别。模型具有较好的识别能力,预警结果的准确率达到了90%以上。

With the greatly ascendance of Chinese netizens for the past few years, the Internet has given the right of free speech to the masses, which making the online-relationship be the topic. The signal of online-relationship is becoming more and more ur-gent. This passage will adopt five calibers pointing to“Xiamen PX incident”,”Kunming PX incident”and“Ningbo PX inci-dent”to have a quantized description. Then, it will forewarn and recognize“Ningbo PX incident”via BP neural international recognition model established by Matlab in order to have a comprehensive study. The model is provided with good recognition ability. Meanwhile the accuracy rate of early-warning reached over 90 percent.