Endless incidents of online violence Technology and law work together to control "buttons" that hurt people

◎Intern reporter Li Zhaoyu

Rectifying online violence and other undesirable online behaviors requires the joint participation of multiple entities such as administrative supervision, internet platforms, and netizens. We should work together to purify the network ecology and create a positive, healthy, clean and orderly network environment.

Chen Lei

Director of the Research Center for Litigation and Judicial System at the Law School of Capital University of Economics and Business

In the past few days, the online violence of the Wenchuan earthquake survivor girl known as "Brother Chunyou" has continued to ferment. Her public private message records show that some netizens verbally insulted and maliciously harassed her.

In recent years, the web has played an increasing role in people's lives. While bringing convenience to people, the Internet has also brought online violence, a "flood beast" that has changed the color of countless people. With the increasing harm of online violence, the control of online violence has attracted more and more attention.

On the morning of March 3, the Information Office of the State Council held a press conference. At the meeting, the relevant person in charge of the Cyberspace Administration of China introduced the relevant situation of the "Qinglang" series of special actions in 28, including the "Qinglang Network Temperament Rectification", a special action related to the governance of online violence.

How harmful is online violence? How to better identify online violence? What should be done to combat online violence? The reporter interviewed relevant experts on the above issues.

The harm of online violence is enormous

"Generally speaking, the behavior of individuals or groups that consciously spread offensive speech through the Internet in order to repeatedly and continuously commit violations against a specific individual or group is an online violence." Chen Lei, director of the Litigation and Judicial System Research Center at the Law School of Capital University of Economics and Business, pointed out that "the forms of online violence mainly include threats, harassment, insults and social isolation, and behaviors are often accompanied by infringement of others' reputations and disclosure of others' privacy. ”

Online violence such as the one that happened to girls who survived the Wenchuan earthquake is not an isolated case. In recent years, the Liu Xuezhou incident and the pink-haired girl incident have all involved online violence. These incidents have caused extremely negative consequences and aroused widespread concern in society.

Some of the victims of online violence either fell into long-term depression and self-doubt, or became victims of serious mental illnesses such as depression, and some even chose to end their lives.

In the face of the harm caused by online violence, it is imperative to scientifically identify and effectively manage online violence.

At a press conference held a few days ago, the relevant person in charge of the Cyberspace Administration of China said that in recent years, the Cyberspace Administration of China has continued to carry out a series of "Qinglang" special actions, cracked down on prominent problems in the online ecology, including online violence, tightened the main responsibility of website platforms, actively responded to the concerns of the people, and effectively safeguarded the legitimate rights and interests of netizens.

Among them, the special action of "Qinglang Internet Temperament Rectification" mainly focuses on strictly managing the information content in the comment area, urging the website platform to strengthen the management of the comment area, and striving to eradicate the breeding ground of prominent problems in the online ecology such as online violence. It is understood that the 2022 "Qinglang" series of special actions carried out in 13 have cleaned up more than 5430.680 million pieces of illegal and adverse information, disposed of more than 2890.26 million accounts, removed more than 7300,<> apps and mini programs, disbanded and closed <>,<> groups, post bars, etc., and closed more than <>,<> websites, achieving certain results in controlling online violence and reducing or even avoiding the harm caused by online violence as much as possible.

Artificial intelligence helps identify online violence

To control online violence, we must first accurately and efficiently identify online violence.

Compared with traditional language violence, online violence is more difficult to identify because of the use of virtualized and obfuscated Internet language, and it is difficult to distinguish it from normal ridicule, jokes, banters and other behaviors. At the same time, due to the characteristics of online communication, the spread of online violence is much faster than that of traditional language violence, and its spread is also far wider than traditional language violence. Therefore, only using traditional manual identification methods to identify online violence is not only time-consuming, laborious, and inefficient, but also prone to misjudgment or even misjudgment due to the subjective reasons of the recognizer, and the overall effect is not satisfactory.

"Public cyberspaces such as social platforms and short video platforms have become places where online violence is at a high rate. Because of its virtualization, zero cost, low cost and other characteristics, the Internet has played a huge role in promoting the generation of online violence, and has also brought great difficulties to the identification of online violence. Yan Huaizhi, director of the Institute of Cyber and Security at Beijing Institute of Technology, said, "Artificial intelligence technology can play a big role in identifying online violence. ”

According to reports, the use of artificial intelligence recognition technology can effectively identify the "source" of various malicious languages, pictures, videos and other online violence in public cyberspace. This advanced algorithm based on artificial intelligence technology has the advantages of more efficient, accurate and objective than manual recognition.

"The biggest role of artificial intelligence technology in curbing online violence lies in its powerful natural language understanding ability." Yan Huaizhi pointed out. For typical unstructured language data such as comments and barrages, artificial intelligence can use technologies such as emotion analysis, text opinion extraction, lexical analysis, and text similarity recognition to achieve full, multi-dimensional, and rapid analysis of online violence, and on this basis, clarify the thoughts, emotions and core views of relevant audiences when viewing related topics, so as to block the links of the transmission chain of online violence and minimize the negative impact and harm caused by online violence.

"It must be pointed out that artificial intelligence recognition technology cannot completely replace manual identification, and the two should be combined to give play to their respective advantages to ensure that the identification of online violence is accurate and reliable." Yan Huaizhi said. If you only rely on artificial intelligence recognition technology, it may cause some ironic identification errors due to its mechanical and fixed defects, and fall into a situation that is too late.

A two-pronged approach to tackling online violence

To effectively stop online violence, in addition to technical means, legal participation is also required.

The network platform can popularize functions such as intelligent filtering of comments, dynamic closure of comments or private messages, identification and blocking of malicious users, etc., to effectively curb the problem of online violence. At the same time, artificial intelligence recognition technology can also be used to remind the offending speakers of their remarks after matching and verifying the big data model of comments, so as to stop online violence in a timely manner.

The reporter learned that in response to the governance of online violence, many domestic network platforms are taking action.

Relevant people from Beijing iQiyi Technology Co., Ltd. told reporters that they have launched the function of private messages to prevent online violence, which can be conveniently used to report online violence with one click and turn off the message function to reduce the negative impact of online violent content on users. The company also actively guides users to post positive content, prompts information about potential online violence, and guides publishers to express goodwill to help create a healthy online environment.

"Perpetrators of online violence may be civil, administrative, and criminally liable depending on how egregious their actions are." Chen Lei pointed out. According to the relevant provisions of China's Civil Code, perpetrators of online violence who cause damage to others shall bear tort liability; If their behavior is more egregious or the circumstances are more serious, they need to bear the corresponding administrative responsibility; If they use physical violence or other extreme methods to carry out online violence, and the circumstances are particularly serious, they need to bear corresponding criminal responsibility.

Chen Lei further pointed out that any individual and organization using the Internet should abide by the Constitution and laws, follow public order and social morality, and must not use the Internet to engage in activities that infringe upon the legitimate rights and interests of others. Rectifying the chaos of online violence and other negative online behavior requires the joint participation of multiple parties such as administrative supervision, internet platforms, and netizens. We should work together to purify the network ecology and create a positive, healthy, clean and orderly network environment. (Science and Technology Daily)