Research On The Early Warning Of Agro-Food Quality And Safety Network Public Opinion

The quality and safety of agro-food are closely related to people's lives.It involves people's three meals a day.In recent years,the quality and safety of agro-food have continued to occur.Not only has the incident caused economic losses,but also the social losses caused by the fermentation of public opinion.The economic loss is hundreds of millions.With the steady development of the Internet industry,the quality and safety of agro-food has caused widespread concern of the government,the public,and the media.In order to timely discover the development status of online public opinion incidents,help government departments respond to public opinion events in a timely manner.This paper proposes to establish a public opinion early warning model for agricultural product quality and safety network opinion,hoping to provide decision support for relevant departments.In order to achieve this goal,this paper first reviews the research status of agricultural product quality and safety network public opinion,and sorts out relevant theoretical research.From the three aspects of the agricultural product network public opinion incident,public opinion media and public participation,three first-level indicators and 11 second-level indicators of agricultural product quality and safety network public opinion early warning index system were constructed,and data sources,indicators meanings and data crawling were constructed.The method is described in detail.Then,through the dynamic comprehensive evaluation sample training method based on gain level excitation,the dynamic comprehensive evaluation value of each public opinion event is calculated,and the training samples are provided for BP neural network,and then the agricultural product quality safety network public opinion early warning model is constructed.Finally,using the Python Keras advanced neural network library for simulation experiments,the accuracy of the test data reached 87%.The results prove that the constructed index system and model are reasonable and feasible.The innovations of this paper are mainly reflected in the following aspects:(1)This paper combines agricultural product quality and safety network public opinion and network public opinion early warning methods to make up for the shortage of agricultural product quality and safety network public opinion warning in quantitative research.(2)It is proposed to use the dynamic comprehensive evaluation method to calculate the comprehensive evaluation value,and the dynamic comprehensive evaluation method can accurately evaluate the time series data.By collecting the information in each time period of the public opinion event,the dynamic evaluation value is calculated,which provides a new idea for the public opinion early warning research.