New media data to boost primary-level social governance

BY LIU JIANGUO and GUO WEI | 09-23-2021
(Chinese Social Sciences Today)

On Apr. 23, a digital intelligence management system was launched in Hangzhou Haiyue Garden Community. The system can automatically monitor, evaluate, and give warnings in such cases that include tracing offenders who throw objects from high-rise buildings in five seconds, alerting car owners who block the fire services in emergencies in two minutes, and watching out for fire incidents 24/7. Photo: CFP


The 14th Five-Year Plan for National Economic and Social Development and Long-Range Objectives for 2035 clearly proposed that the social development goals in the new era should include: “constructing a new pattern of primary-level social governance,” “improving the social governance systems featuring co-construction, co-governance, and sharing,” and “strengthening and innovating social governance.”

 
At the same time, the popularity of smart phones and other modern communication tools, as well as the rapid development of new media such as Weibo and WeChat, have not only facilitated people’s participation in social governance, but also brought about new ideas and methods for it.
 
Reflection of public opinion 
Social governance is an extremely complex “super project.” In recent years, the rapid advancement of cloud computing, big data, and other new technologies has provided strong support for the intellectualization of social governance, shortening the time from problem detection to emergency response, and thus improving efficiency. 
 
However, with the popularization and application of big data in social governance, some problems have also arisen. For example, all kinds of objective data generated by smart devices can be used to analyze the objective behaviors of urban residents, so as to achieve a reasonable allocation of resources. In the process, however, there is no way of knowing if residents can achieve the “sense of gain” from the effect of community-level social governance. As a result, it can be said that the closed-loop management of the “perception-prediction-response-effect evaluation” system has yet to be formed in social governance. 
 
At present, the construction, configuration, and maintenance of public facilities can be monitored and managed in a timely manner through big data technology and means. However, it is difficult to learn “tacit knowledge” such as “sense of happiness” and “sense of gain” in social governance at the community level. Given that relevant information on primary-level social governance relies heavily on communities and government service hotlines, which cannot cover all residents and are troubled with imperfect information and inaccurate expression, it is impossible to learn people’s differentiated appeals. This poses potential risks to social governance in the long run.  
 
In this light, new media data is intended to collect and organize information obtained from social media apps, the public affairs Weibo, WeChat public accounts, online forums, Tieba, and short-video platforms, which can help government departments at various levels to detect community-level social governance problems in real time, make rapid responses, and promptly evaluate the effect of the targeted management. 
 
That said, new media data has yet to play a big role in primary-level social governance, due to the following aspects. First, in most cases, only major events that have attracted public attention are caught by the social governance radar, while somewhat overlooking people’s appeals and demands in daily life. Second, keyword monitoring is often deployed in cases of emergencies, which sometimes leads to omission of information and untimely detection. Third, new media data is not incorporated into the social governance information system at the moment, so it fails to serve different government departments in a holistic manner or form a closed-loop management system of “perception-prediction-response-effect evaluation.” 
 
Multiple functions
In the future, new media data needs to assume a more important role in primary-level social governance in terms of reflecting public opinion. On top of this, artificial intelligence technology and regional governance evaluation index systems can be combined to design an emergency perception system, risk prevention system, major event tracking system, and other subsystems, so as to carry out early warning, situation awareness, and tracking of on-going emergencies. In the end, the good integration of internal government data and new media data is expected to better empower digital governance. 
 
To start with, it is essential in social governance to assess whether the public’s concerns can be sensed at the infancy of a problem through the situation awareness of all media, so that the problems can be solved in advance before they escalate into crises. At the current stage, we believe government agencies at all levels can better facilitate technological advancement to detect potential social governance risks early on through all-media data, and then intervene in time to resolve and eliminate many potential social conflicts in advance.
 
The first step of primary-level social governance is to extract people’s appeals and demands, and then use big data analysis and mining technology to identify the relevant subjects involved in emergencies, so as to predict and deduce the evolution dynamics of the events. The second step is to look for patterns in all the little things that new media data reveal. By analyzing the distribution of regions, areas, government departments involved, and the intensity of discussion, we can find the principal contradictions and the topics of greatest concern among the public. The third step is to superimpose the emotional distribution of people on the internet with these events, and apply it to analyze the existence of potential risks.
 
At present, due to the high popularity of mobile internet, as long as there are netizens at the scene of an emergency, in most cases, there will be photos or videos uploaded to the internet instantly. This kind of information can be used as a good supplement to information sources such as internal government alarms and complaint hotlines in emergencies. 
 
What’s more, it is necessary to develop emergency awareness, which is different from keyword monitoring that relies on manual selection of information in large flows. The system is based on artificial intelligence technology, and can automatically locate events on its own initiative from real-time information sources such as Weibo, BBS, and Tieba. With the aid of various kinds of emergency data models, we can carry out spatial-temporal calculations and accurately spot the emergency in a timely manner. 
 
Such a system can quickly find problems with potential risks in different urban areas, and issue early warnings based on the intensity of online discussion. At present, it has been used in the perception of emergencies in fire protection, people’s livelihoods, education, medical care, disasters, transportation, municipal administration, and other aspects.
 
Early warnings and situation awareness can help government departments better find and resolve emergencies in primary-level social governance. After the occurrence of an emergency, the focal point is to “solve the problem of people,” meaning that we cannot overlook the accident site, and the help needed by those involved in the emergency, as well as the public’s reaction and perception. Therefore, the emergency tracking and analysis system can help relevant departments to make more detailed and targeted policies.
 
Expectations
If we want new media data to play a bigger role in primary-level social governance, the following three problems need to be better addressed.
 
New media data excels at intuitively revealing the social atmosphere and public opinion. However, due to the difficulty of data acquisition and analysis, it has yet to be widely used in areas other than the field of public opinion. Most of the new media data and complaint data are unstructured data, huge in quantity and complex in nature, leaving it far more difficult to apply than structured internal government data. Breaking this bottleneck will require the use of a natural semantic analysis ability, a complex computing ability, and a massive new media data real-time analysis ability.
 
An outstanding feature of new media data is that it comes from the open internet platform, free from the restrictions of departments, regions, time, and channels. It is instant and authentic, and does not require manual layer screening. Any agencies that are technically equipped can take advantage of the data.
 
At the same time, it can cooperate with smart city management departments to explore more and deeper application scenarios. Through data complementarity, data comparison, and fusion presentation with internal government data, new media data may be able to help form a perfect emergency detection and management procedure. Through the combination of internal and external data, timeliness, authenticity, and comprehensiveness of information in primary-level social governance can be enhanced.
 
Third, it is significant to gain a timely and comprehensive perception of public opinion in emergencies. More and more real-life cases remind us that it is far from enough to only monitor and track public opinion. It is necessary to grasp rapidly changing public opinion, understand the distribution of public sentiment and its causes, and perceive potential risks from abnormal trivial events in primary-level social governance.
 
Liu Jianguo is a professor from the School of Accountancy at Shanghai University of Finance and Economics and Guo Wei is from Shanghai Pudong Hot Topics Big Data Research Center. 
 
 
 
 
Edited by YANG XUE