The arrival of the era of big data has had a revolutionary impact on human life, work and thinking, and profoundly changed the face of the business kingdom and public management. "Big data" has increasingly become a booster for innovation in various industries. At present, China's network public opinion environment is complex, and network public opinion crisis occurs from time to time. Social hot public opinion events and official public opinion events are constantly emerging, which has caused many influences such as the imbalance between social democratic life and political stability. The online public opinion under the background of big data is undergoing tremendous changes, and the management of online public opinion has become increasingly complex and important. How to seize the opportunity brought by the era of big data, change the traditional thinking of online public opinion management with the concept of big data, accurately grasp the inherent characteristics of online public opinion and its potential laws in the evolution process, and realize the innovation of online public opinion management in thinking, mode and technology is of great theoretical significance for guiding online public opinion and strengthening and improving online content construction under the new situation.
First, the era of big data will inevitably require changes in online public opinion management.
The concept of "big data" was first put forward in the 1980s. On 20 1 1 year, McKinsey & Company released its research achievement "Big Data: The Next Frontier of Innovation, Competition and Productivity", which made this concept widely popularized. On March 29th, 20 12, Obama announced that he would invest more than 200 million US dollars to launch the "Big Data Research and Development Initiative" and upgrade the "Big Data Strategy" to a national strategy. In the past two years, big data has attracted the attention of academia, industry and government departments and has become a powerful frontier vocabulary at home and abroad. Big data, also known as massive data, massive data and big data, refers to an excessive collection of data captured, managed and processed by mainstream software tools within a reasonable time. It is a huge amount of information, and only through deep mining, calculation and analysis can we create value. Big data greatly surpasses the traditional data form in terms of volume, complexity, production speed and value density, and has 4V characteristics: volume, type, speed and value. A large number of netizens express their opinions conveniently and quickly through forums, Weibo, WeChat and other channels. The scale and complexity of online public opinion are rising rapidly, with huge volume and low value density. The change of its inherent characteristics will inevitably require the transformation of network public opinion management to adapt to the development of the era of big data. These requirements are mainly reflected in the four "turns".
From monitoring to forecasting. The core and goal of big data is prediction. Barbara Brazil, an expert on complex networks, believes that "93% of human behavior is predictable. When we digitize, formulate and model our lives, we will find that everyone is very similar. Life is so resistant to random movement that it is eager to develop in a safer and more regular direction. Human behavior seems to be very random and accidental, but it is extremely easy to predict. " [1]. For example, Amazon can recommend books we want, Taobao knows our preferences, and Renren can guess who we know. The traditional network public opinion management takes the public opinion information generated by monitoring as the starting point, and this obvious lag makes it in a passive position in dealing with the network public opinion crisis. At present, the time left for dealing with emergencies is getting less and less, from the traditional "golden 24 hours" to "golden 4 hours". In such a short time, public opinion analysis and decision-making have not had time to participate, and the whole incident has already caused an explosion effect. In the era of big data, by mining the correlation of data, the mathematical algorithm is applied to massive data for analysis, sensitive news is monitored in advance in the initial stage of network communication, and then a model is established to simulate the evolution process of network public opinion, thus predicting the possibility and tendency of network public opinion emergencies.
(2) From node to network. From monitoring public opinion to predicting public opinion, the most critical big data technology is to mine the correlation of data. In the era of small data, due to the limitation of database and computational analysis ability, it is time-consuming to pursue causality or correlation, and it is easily influenced by traditional thinking mode and inherent bias implied in specific fields, so the accuracy of public opinion analysis results cannot be guaranteed. Therefore, the traditional network public opinion management only pays attention to the monitoring of public opinion content, and grasps the shallow social semantic expression by analyzing individual data nodes, such as what netizens said. While retaining the original data, big data records the netizens' "Why do you say that?" According to the big data thinking, each data is a node, which can infinitely form a multiplication effect with other related data in the public opinion chain-similar to the Weibo fission propagation path, and the fission-related state of data contains infinite possibilities [2]. Through the deconstruction and reconstruction of massive information, the data assets of government and enterprises are fully integrated, and a series of rapidly developing new technologies and tools are used to describe, measure and calculate the relationship between nodes, so as to deeply mine the correlation of data, thus eliminating prejudice and visual blind spots, grasping the social trends that are easily overlooked and predicting the development trend of public opinion. Therefore, in the era of big data, network public opinion management must change its monitoring system, from nodes to networks, grasp the relevance, and then analyze the social interaction behind public opinion, and even the boundaries and interconnections between network ethnic groups.
(3) From qualitative to quantitative. From their own experience and perspective, when public opinion analysts or interpreters conduct qualitative analysis in the process of traditional online public opinion management, their analysis results will inevitably bear the subjective imprint of personal values and ideas, and even different public opinion organizations will draw contradictory conclusions about the same public opinion event. In the era of big data, all metadata can be transformed into valuable information through quantitative association and can be used many times. Every use is an innovation, and big data has become the source of strength for quantitative management of online public opinion. Although the correlation of data determines the potential value of some data, the emergence of new technologies and new software makes it possible to realize the value transformation of data through mathematical analysis. The multi-dimensional interpretation of public opinion and the revelation of new profound insights make the results of public opinion analysis comprehensive and objective, which greatly exceeds the traditional network public opinion management. However, the quantification of data is not the same as simple "digitization", but the computability of data, which Schoenberg calls "digitization", which refers to the process of transforming phenomena into quantifiable forms and making tables and analysis [3]. "Digitalization" transforms attitudes and emotions into analyzable forms, which can deeply analyze the relevant information of online public opinion. Some social media, such as Facebook, Twitter, QQ, Weibo, WeChat, etc., are sitting on the treasure of big data. Once their databases are deeply used, they can easily obtain all the dynamic information of almost all social fields and all users.
(4) From the sample to the whole. In the traditional network public opinion work mode, the collected public opinion related data is only sample information, and the database structure is single and the data volume is limited. Its data source is generally based on sampling or data grabbing of key network sites, and it can only analyze small-scale, structured or quasi-structured data, with different standards, so it is difficult to use in different fields. At the same time, sample analysis cannot guarantee the accuracy of the results. Even if there is no problem with the analysis method and operation, any mistakes in the sampling process will make the results of public opinion analysis far from the facts. Big data has a huge volume, jumping from TB level to PB or even ZB level, completely recording social conditions and public opinion, and becoming a recorder of human survival traces and psychological changes. The purpose of sampling is to get as much information as possible with as little data as possible, but big data is based on mastering all data, at least massive data. With the rapid development of data processing technology, it is necessary to change the traditional thinking and methods of public opinion management and the inertia of sampling. Using big data technology, an automatic analysis system of online public opinion is established, which can automatically search and collect information that seems to have nothing to do with the target public opinion, but is internally related. After grabbing the collection page, it automatically classifies information, automatically obtains keywords, and automatically analyzes the content and gives an alarm. The sample is expanded to almost all, and the results of public opinion analysis are more objective and reliable.
Second, the effect prospect of online public opinion management reform in the era of big data
Seize the new opportunities to change network public opinion management in the era of big data, meet the new challenges and meet the new requirements of network public opinion management in the era of big data. Changing and innovating network public opinion management will produce good management effect and realize the upgrading and transformation of network public opinion management in the new period.
(A) to achieve "fire prevention" management. Because the traditional network public opinion management can't grasp the data correlation and accurately predict the future development trend of public opinion, it adopts the "fire-fighting" management mode. Usually, the government will only take measures when public opinion is generated or has formed a public opinion crisis, such as releasing information, guiding public opinion and satisfying demands. , so as to achieve the effect of "fire fighting". In this mode, the government often passively falls into the whirlpool of online public opinion, forming a prejudice that online public opinion is regarded as "the enemy's situation". In order to get rid of this dilemma, the government always tries to "control", "guide" and "respond" to online public opinion, and dominate and dominate netizens and their public opinion expression with a superior attitude. However, if the dominant position of netizens in online public opinion is not guaranteed, online public opinion will lose its function as a "pressure reducing valve", and the problem of online public opinion will take the temporary solution rather than the root cause. In the era of big data, the government will change the thinking and mode of network public opinion management, and apply big data technology to carry out correlation analysis, grade division, cluster analysis and trend analysis on network public opinion, so as to realize the transformation from "fire fighting" management to "fire prevention" management. Only by finding the balance between "fuse" and "pressure reducing valve" can we play the role of network "public opinion field" and stifle the network public opinion crisis in the cradle. For example, the CIA tracked terrorists and monitored social sentiment by capturing massive data. In the Arab Spring, it analyzed how many people and whose positions changed from moderate to radical through big data, and "figured out" who might take harmful actions.
(2) salvage "heavy sound". Big data stems from the sharing and openness of the Internet, but the existence of the "digital divide" isolates the "information poor" from the network. Although the development of the Internet makes the proportion of this group of people lower and lower, the expansion of uneven development means that there is another group that cannot be ignored and will not be able to provide any data. Even people who can make full use of the Internet may become vulnerable groups in public opinion in some cases, or choose not to speak on the Internet because of heterogeneous thinking in the mainstream of public opinion. Of course, this choice can be active or passive. As the American philosopher eric hoffer said, "The most inactive people in a country are the middle class. They are decent people, working in cities and farming in rural areas. However, their fate is dominated by a few people at both ends of the social spectrum-the best people and the worst people. " [4]. Obviously, the big data platform built by the technical system alone cannot really obtain "all data". It is necessary to salvage those "sinking voices" that may represent a certain group or a certain order of magnitude by reforming the management of online public opinion. Therefore, it will be helpful to salvage the "sinking voice" by comprehensively thinking about and clarifying the opportunities and challenges faced by network public opinion management in the era of big data, and changing the working concept and mode of network public opinion management through the construction of the concept of "big public opinion". For example, combining public opinion service with social investigation, paying attention to on-the-spot investigation and first-hand material collection, rather than bundling online public opinion management with technology, will avoid getting incomplete public opinion or making misleading decisions.
(3) See through "false public opinion". At present, the network public opinion, which has attracted much attention, has increasingly become a "pseudo-public opinion" centered on movie star scholars, movie star journalists, movie star businessmen and movie star politicians [5]. After the occurrence of major sensitive events, some network managers and influential public opinion organizations quickly shielded their subjective "harmful information" and selectively compiled public opinion reports to influence decision makers' judgment of the situation with one-sided and false "false public opinion" and make them make decisions in line with their own interests. Some interest groups carefully foster and cultivate their own network spokespersons to guide netizens to think about the content and direction. As a result, these opinion leaders' views on key events and issues are popular on the Internet, and other heterogeneous opinions are submerged, which makes people's cognition of the truth greatly deviated. When public opinion is manipulated by the political and economic forces of various interest groups, it loses its independence. Once "pseudo-public opinion" is discovered, public opinion institutions may lose their credibility. The complete, accurate and fast information capture based on the whole network is conducive to providing first-hand materials and pure facts for the public opinion analysis report, so as to obtain real and comprehensive public opinion, so that netizens can still understand "what" fairly and objectively without knowing "why", thus helping to guide online public opinion. At the same time, by changing the system and mechanism of online public opinion management, we can maintain the independence of public opinion management, effectively identify "false public opinion" and eliminate "noise" and "noise", so that online public opinion in the era of big data can truly become a "mirror image" of the real world.
(4) Overcome "the blind touch the elephant" and "the information island". The contradiction between the infinite growth of massive information and the limited attention and analysis ability of netizens has caused the strange phenomena of "data explosion" and "lack of knowledge", which has intensified the "blind people touch the elephant" effect of public opinion. In the era of big data, online media has promoted the openness of information and the convenience of dissemination, and people's participation in public events has reached an unprecedented height. However, the prominence of focused communication and personalized communication and the fragmentation of information make it more and more difficult to pay comprehensive and profound attention to and analyze the incident. The irrational and excitable characteristics of netizens lead to the extreme and emotional network public opinion, and the "group polarization" of the network is amplified. Public opinion monitoring in the era of big data is based on the collection of public opinion information on the whole network that traditional manual and software can't do, and the samples are expanded to all. Using big data technology, an automatic analysis system of online public opinion is established to avoid the lack of important information monitoring caused by incomplete data sources, which is helpful to eliminate the phenomenon of "blind people touching the image". At the same time, due to the uneven level of informatization application, there is an "information island" problem between different government departments and enterprises: there are as many information systems as there are departments, and each system has its own database, application software and user interface, which is completely independent, which hinders the interconnection of data [6]. Changing the working mode of online public opinion management in the era of big data, unifying the technical standards of public opinion industry, realizing data sharing, establishing an online public opinion service alliance, coordinating the government, enterprises, media and social forces, and realizing the pluralistic governance of online public opinion will help solve the problem of "information island".
Third, the transformation path of network public opinion management in the era of big data
When big data has revolutionized all walks of life, the whole world is not ready for this industrial revolution. But compared with developed countries such as Britain and America, China is more like the eve of the era of big data. China's population and economic scale determine that the scale of big data in China is the largest in the world, which provides a rare opportunity for China to seize the pulse of the times and carry out reform. In this context, big data has also had a profound impact on traditional public opinion management. In order to make the network public opinion management reform achieve the expected effect and adapt to the requirements of the times, we must start from the aspects of thinking concept, methods and means, institutional mechanism, technical support and talent construction.
(1) Establish the concept of public opinion. The transformation of network public opinion management in the era of big data lies in establishing the concept of big public opinion. The big public opinion here includes two meanings. First, emphasize the "big data view", that is, fully realize the open sharing of network data platforms. According to the big data logic of "everything can be quantified", the generation of a new associated data usually brings a new analysis result. Therefore, only by forming a "big data view" and realizing the dynamic sharing of data can we effectively prevent information fragmentation and eliminate the phenomena of "blind people touching the elephant" and "information islands" to the maximum extent. Second, emphasize the integration of online and offline data. The insufficient combination of online public opinion and social survey may reduce the authenticity of public opinion and mislead decision-making. For example, for the choice of holiday adjustment scheme, the online voting results organized by public opinion organizations are different, and the public opinion analysis report they made is also different from the real public opinion. Therefore, only by truly grasping the "great public opinion" and salvaging the "sinking voice" can we make correct decisions and build a safer and more efficient society. To establish the concept of public opinion, we must first realize the dynamic analysis of data, break the data monopoly, unify standards, share data, and prevent isolated public opinion organizations from making one-sided or wrong public opinion analysis reports behind closed doors. Secondly, it is necessary to integrate online and offline data, tap the deep relationship between online public opinion and social dynamics, and realize the close linkage and synchronous promotion of online public opinion management and social governance [7]. Finally, improve and innovate all aspects of online public opinion management, including public opinion capture, early warning, judgment, decision-making and evaluation, so that the public opinion management function is not limited to crisis handling, but also plays a role in assisting decision-making.
(2) Change the guiding strategy of online public opinion. To do a good job in guiding public opinion, we must grasp timeliness, appropriateness and effectiveness. However, at present, many localities and departments still lack a correct understanding of how to guide online public opinion. They can't grasp the "golden four hours" in "time" and the "temperature" and "efficiency" to ensure the actual quality of online public opinion guidance. Because of its own characteristics, big data is conducive to changing the guidance strategy of online public opinion, changing "blocking" and "ostrich tactics" into "online guidance, where to fall online", making "pseudo-public opinion" lose its living soil. Therefore, we should give full play to the advantages of big data and improve the ability of public opinion guidance. First, use big data to improve the predictability and purpose of online public opinion guidance. Through data capture and correlation analysis, we will build an analysis model of netizens' opinion tendency, understand netizens' preferences and characteristics, build and improve government websites and official Weibo, and cultivate and rely on opinion leaders who can talk, talk, be grounded and do practical things. The second is to realize the value guidance of online public opinion through the value transformation of data. On the basis of fully collecting relevant data, data visualization techniques such as charts are used to reveal the cause and effect of the incident, so that the data can be "voiced" and netizens can "know what it is" and "know why it is", so as to understand the cause and effect of the incident in a 360-degree way and put an end to the phenomenon of "blind people touching the elephant". Third, enhance the credibility of public opinion guidance. On the one hand, strengthen the interaction between new and old media, give full play to their respective advantages to communicate with the public, crack rumors and realize the double guarantee of timeliness and authority; On the other hand, it is necessary to prevent public opinion analysts from being influenced by experience preferences in the process of processing data, and to prevent big data from becoming a means for some institutions and individuals to manipulate public opinion more conveniently.
(3) Improve the system and mechanism of big data public opinion management. At present, the system and mechanism of online public opinion management is not perfect, and there is no systematic and standardized public opinion response and handling management system in many areas. Problems such as backward means of public opinion analysis and prediction, lack of crisis response system, imperfect and unstable public opinion management institutions, and multi-head management are very common. Perfecting the system and mechanism of big data public opinion management plays a decisive role in solving the problems and difficulties in the process of online public opinion management from the source and realizing both the symptoms and the root causes. Therefore, in order to make online public opinion management effective and improve the standardization and scientific level of online public opinion work, China should speed up the establishment and improvement of the system and mechanism of big data public opinion management. First of all, we should establish an interactive mechanism for multi-management of online public opinion. The state will issue a strategic plan for the development of big data, and Industry-University-Research will coordinate the efforts of the government, enterprises, society and citizens to form a joint force to achieve co-governance. Secondly, change the institutional setup of online public opinion management, change the previous mode of passively responding to public opinion crisis by temporarily setting up a leading group or a temporary office, or taking the propaganda department as a "fire brigade", and make online public opinion management professional and refined through normalized institutional setup and professional personnel. Thirdly, establish a clear responsibility mechanism, and clarify the rights and obligations of all levels and departments, including government departments, corporate media and people's organizations, by accelerating the process of data legislation; By establishing a big data public opinion management system led by the network information department, the situation of multi-head management will be changed and the government chief information officer responsibility system will be established. Finally, improve the resource guarantee mechanism of public opinion management in big data network. In the era of big data, the transformation of online public opinion management is faced with problems such as high initial cost and insignificant short-term benefits, and it is necessary to increase investment in resources such as capital, technology, materials and manpower.
(4) Innovating the methods and technologies of public opinion management in big data network. With the advent of the era of big data, it is required that online public opinion management must adopt more advanced technologies, which are mainly manifested in the wide application of various related software and the supporting platform of big data technology. At present, the representative China public opinion monitoring collection software includes TRS Internet public opinion information monitoring system, Peking University Founder Zhisi public opinion monitoring system, military dog network public opinion monitoring system, and Lesi network public opinion monitoring system. In addition, we should improve and innovate the five cornerstones of the big data technology support platform-data monitoring technology, data mining technology, data storage technology, data analysis technology and data security technology, so that big data can serve the network public opinion management without exceeding our control. At the same time, we can't fall into the misunderstanding of "technology is omnipotent" and blindly rely on technology, and we can't lead to "data dictatorship" and become a slave to data because we believe in the powerful forecasting function of big data. Therefore, network public opinion management also needs to rely on other methods and means to complement each other and work together. Law has become the most effective means of management and control because of its greatest compulsion and authority. Law and morality are interrelated, and education and self-discipline are placed in an important position in the virtual space with great complexity and particularity. For example, developed countries in Europe and America, such as the United States, Britain, Canada and so on. All advocate user self-discipline and self-management to improve the media literacy of netizens and strengthen their self-control ability. In addition, we can follow the example of South Korea and Singapore and use administrative means to require network users to obtain the license issued by the relevant state departments before accessing the information strictly controlled by the government.
(5) Cultivate network public opinion management talents in the era of big data. Internet public opinion in the era of big data will form multi-dimensional research, such as the analysis and research on social discourse expression, social psychological description, social relationship presentation, social appeal prediction and so on. Internet public opinion will truly become a multidisciplinary social science, which requires high comprehensiveness of talents. The discipline division and training system of education in China objectively make it difficult for the trained talents to cross the border. In other words, the threshold for truly entering this industry is very high. For this reason, countries pay more and more attention to the training of data scientists. For example, the United States has set up a special course in universities to study big data technology, and trained the next generation of data scientists through strict business training and professional qualification certification. On September 20 13, Ministry of Human Resources and Social Security and People's Daily launched the "Network Public Opinion Analyst Vocational Training Program", and "Network Public Opinion Analyst" became an officially recognized profession. However, the level of existing public opinion workers in China is still seriously lagging behind, and many public opinion organizations, especially local governments, do not have professional data processing and analysis teams and specialized online public opinion management departments. In order to break through the talent bottleneck of changing online public opinion management in the era of big data, we can introduce data mining and analysis talents through recruitment and employment in the short term, strengthen the strength of existing professionals through entrusted training and online training, and rent high-quality talents for big data public opinion management in the short term by purchasing services. In the long run, it is necessary to systematically sort out the list of talents needed for network public opinion management, cultivate and expand compound talents who are proficient in data mining and mathematical modeling, have high learning ability, analytical ability and knowledge level, and span disciplines such as statistics, sociology, computer science, communication, management, etc., and build a talent team for big data network public opinion management.
The above is Bian Xiao's discussion on the transformation of network public opinion management in the era of big data. For more information, you can pay attention to Global Ivy and share more dry goods.