Design And Implementation Of Public Opinion Monitoring And Analysis System For Special Equipment Accident And Fault Event

With today’s developed network,the Internet has become the most important,the most massive information carrier.The public opinion information on the Internet has a profound impact on society and people.Therefore,the supervision and mining analysis of the public opinion information on the Internet has great value to the government and enterprises.The Quality Supervision Bureau,which is responsible for the supervision of special equipment safety,also tends to monitor the public opinion information of important special equipment.Special equipment is some of the more dangerous equipment,such as elevators,gas cylinders.Therefore,the influence of its failures or accidents is relatively large,and public opinion information about that will be widely disseminated on the internet.Therefore,the monitoring of the related public opinion of the special equipment accident or failure event can help the supervisor to give timely feedback and disposal of the related events.Based on the requirement of monitoring and analyzing the public opinion of special equipment’s accidents and failures,this paper studies the related technologies of public opinion monitoring and analysis,and proposes improvement methods for the core algorithm part.Based on the improved algorithm,this paper designs a monitoring and analysis system for the public opinion of special equipment accident and fault events.This paper first introduces the background of public opinion monitoring and analysis,and it gives a brief introduction to the monitoring and analysis process of public opinion,the technology involved in the crawler and the natural language technology also are introduced.Then the requirements of the public opinion monitoring and analysis system for special equipment accidents and failures are described and analyzed.Then,based on the requirements of monitoring and analysis of public opinion in special equipment failure and accident event,two core algorithm modules of public opinion monitoring and public opinion clustering are discussed and optimized.In the part of related public opinion identification,a recognition strategy combining keyword and machine learning classifier is proposed.In the clustering part of public opinion events,a hybrid feature based similarity calculation method for public opinion text is proposed to optimize the clustering results.Finally,a public opinion monitoring and analysis system for special equipment accidents and faults is designed,based on the optimized algorithm.And the experimental evaluation to the relevant public opinion recognition and public opinion clustering algorithm are carried on,with the public opinion data collected by the system.Experiments show that the relevant public opinion recognition and public opinion clustering algorithm proposed by this paper both perform better than classic method to some extend.