Fundamentally speaking, network public opinion monitoring requires real-time, comprehensiveness and accuracy. Duoruike public opinion data analysis station can meet the three major elements of the system public opinion information service market, namely, real-time: the earlier the public opinion is obtained, the greater the space left for response and disposal, while the development of online public opinion is often explosive, and the opportunity is fleeting. Once the best opportunity is missed, it will be regretted; Comprehensiveness: Internet public opinion is a little easier to attack, not as good as the rest, with a large amount of information, and it is easy to ignore many important information; Accuracy: under the condition of massive information, how to accurately grasp the core of the problem and not be trapped by interference is the key to making a correct judgment.

At present, public opinion monitoring has a lot to improve. First of all, manual monitoring has natural limitations. By arranging fixed personnel to be on duty 24 hours a day, constantly browsing target sites and searching for target keywords, it is the most direct and primary public opinion monitoring method in the absence of automation system. Due to the limitation of everyone's subjective thoughts, there will always be blind spots in manual monitoring, and there will always be places that are unimportant but proved to be serious afterwards. Manual monitoring cannot detect the changes of some websites or some remote web pages; At the same time, people are not machines, and repeated monitoring for a long time will easily lead to fatigue, and often the judged public opinion will be missed if you are not careful. These will fluctuate greatly in real-time and accuracy.

Second, over-reliance on search engines. People often think that only search engines have the ability to find a needle in the ocean of online public opinion. However, search engines still have many limitations. Except that the search results are greatly influenced by keywords, the results returned by search engines often come from websites that are not subject to any access restrictions, while websites that need to log in, such as forums, are completely excluded. However, it is precisely these websites that provide interactive functions that netizens express their opinions the most. Moreover, the web crawler of the search engine has a certain time delay, so it can't search for the latest network updates in real time. So the search engine can't provide us with the whole picture of the problem, because it only pays attention to keywords and doesn't pay attention to the problem, so it can't be comprehensive; Search engines can't get the results we want in the first time, because not only its content is not updated quickly enough, but also it can't provide timely services, so it can't be timely; Search engines can only return another massive result in a massive web page, and it is in a chaotic form. It knows nothing about what we care about, so the result it gives is what everyone cares about, not what I care about, so it is not accurate. It can be seen that we can't rely entirely on search engines to monitor online public opinion, and we need to propose new technical means to learn online public opinion and track it.

Therefore, in order to ensure the real-time, comprehensiveness and accuracy of public opinion monitoring, the most reliable way is to use an automated public opinion monitoring system, rely on software systems to eliminate the shortcomings of manual methods, rely on software systems to monitor target sites and the whole network at fixed points, track and analyze the development tracks of various public opinion topics, and automatically sort out and generate daily/weekly reports, thus organically combining public opinion with the daily work and business of the government.

Dorek public opinion data analysis station system is an application software system based on Internet public opinion information management. The system integrates Internet information collection technology and information intelligent processing technology, and realizes users' information needs such as public opinion monitoring and news topic tracking on the Internet by means of automatic crawling, automatic classification and clustering, topic detection and topic focusing, and forms analysis results such as briefings, reports and charts, so as to provide customers with a comprehensive grasp of the ideological trends of the masses, make correct public opinion guidance and provide analysis basis.

The common method of negative news on the Internet is to delete posts, but sometimes deleting posts can not only not ease public opinion and calm it down, but will further intensify public opinion and public dissatisfaction, further ferment and escalate the incident, and evolve into a public * * * event and a public * * * topic discussed by the whole society, which is not conducive to the handling and solution of the problem, leaving the government and enterprises in a more passive situation, resulting in damage to the credibility of the government and enterprises.

So how to deal with the negative news of the network?

First, improve network awareness and avoid topics that lead to negative hype. This is prevention.

The second is to integrate internal institutions and build a rapid response public opinion monitoring system. We can establish a public opinion monitoring system, such as ant workshop software, to find the bud of negative news on the internet as soon as possible. This is surveillance.

The third is to dispel rumors, channel and block simultaneously, and firmly grasp the right to speak through online public opinion. This is guidance.

The fourth is to expand all kinds of voice channels and establish and improve the network news spokesperson system. This is post-processing.