Big data prose

▶ The national procuratorial organs should use the "class case" as the starting point to create a digital "case supervision" as the core task of empowering legal supervision by big data.

▶ Compared with case supervision, the supervision of cases is a high -level state that cannot be split and has a significant improvement. The technique of the procuratorial organs to carry out case supervision is that in response to the common problems found in the case, the accurate and efficient supervision that promoted the solution of such problems, and finally achieved the social governance of "handling a case, supervising a batch, and governance". efficacy.

Renmin University of China

Professor and doctoral supervisor of law school

Researcher at Criminal Law Science Research Center

Liu Pinxin

The "centered on the supervision of cases" is one of the essentials of the Supreme People's Procuratorate's clear development of digital procuratorial innovation. Specifically, the national procuratorial organs should use the "class case" as the starting point to create a digital "case supervision" as the core task of empowering legal supervision by big data. "Cases" are the keywords, and it is self -evident to explore the important value of "big data+procuratorial supervision" (hereinafter referred to as "big data procuratorial").

The so -called "case" is the synonym of "the same case" in the principle of "the same case and the judgment". For a long time, the people have a simple favor about the "similar cases and the same judgment". The academic circles have made theory expressions such as "similar cases, similar judgments", "similar cases, similar judgments", "similar cases, the same judgment", "the same circumstances, and handling" and other theories. These requirements are based on the perspective of the case, focusing on the continuity between case processing. Since the emergence of smart technology to embrace judicial exploration, the theory of new "class cases" with the same elements of legal documents, similar texts or provisions, etc., has enriched the existing concepts. Correspondingly, new events such as "Case Retrieval", "Classification of Cases", and "Class Classification" based on case -based cases have emerged endlessly. basic work. The reason is that the data containing the elements of the case is "learned" by the machine to form a machine wisdom that is different from human wisdom. This shows that the case is no longer only as a ruler of judicial justice, but also becomes the source of machine learning in innovation of "law+technology".

For the exploration of big data procuratorials, the phenomenon and laws of class cases cannot be ignored, and the principle of relevant cases can also be nested and re -created. The key is how to understand the basic issues of big data procuratorial from the perspective of the case.

What is the big data?

In short, procuratorial big data refers to the amount of data that reaches a certain level and can be used to discover various electronic materials that require the characteristics of the case or laws required for procuratorial cases. Scanning, optical characters recognition, and recording of traditional media such as paper dumplings such as paper dumplings are also listed here. Of course, the amount of data of procuratorial big data is difficult to quantify. From the perspective of experience, the amount of data that transcends the manual analysis capabilities of the prosecutor or the case handling team can be regarded as procuratorial big data. For example, the electronic data of the CD -ROM and hard disk in a criminal case is the data treasure at hand at the prosecutor. From a technical perspective, of course, the amount of data is better and pure, otherwise it is difficult to show the advantages of use.

In comparison, whether the characteristics or laws that can be found in the data from the data to discover the cases needed to handle the case are the real standards for judging the effectiveness of procuratorial big data. This requires the data used by prosecutors to reach a low -limit order. The low limit of procuratorial big data is not high in specific scenarios. The author's team has done some related experiments, such as using a multi -linear regression modeling method to train the crime of destruction of public telecommunications facilities to generate a sentencing prediction model to supervise abnormalities. At that time, it was found that once the referee documents for training data reached 20,000 copies, the accuracy of the sentencing forecast model was relatively stable. In other crimes related trials, we have also encountered less than 20,000 legal documents. In this case, a certain degree of remedy can be carried out, such as multiple use of the same batch of training data to play the role of data amplification.

The procuratorial big data that meets the requirements of low limit can be used to support digital procuratorial innovation. The mystery is to refine the characteristics or rules of cases or rules through seeking same laws and differences. If the analysis has obtained abnormal signs that do not meet the normal laws in similar cases, you can be used as a clue of legal supervision. To complete these tasks, we usually need to use technical models such as data portrait, data collision, data mining, etc., and will be further expanded below.

As for the specific state of procuratorial big data, legal documents, data applicable data, evidence admission data, execution data, people's livelihood data, clue research and judgment data, resource protection data, lawsuit abnormal data, non -litigant data, non -litigant data, resource protection data, resource protection data, database data Data, administrative non -appealing data, online public opinion data, judicial network photography data, insurance claims data, supplementary (rescue) fund supervision data, criminal records of sealing data, industry compliance data, etc. are all in the procuratorial big data. It also supports the development of the current legal supervision model of the procuratorial organs. In fact, electronic resources such as academic papers, digital archives, conferences, and professional websites in the field of law research can also be used for digital procuratorial supervision and practice. The author has investigated and learned that when a local procuratorial organs investigate and deal with the case of a medical unification, the author's retrieval function of the journal papers database is used for correlation analysis, and the medical workers involved in the case are convenient to distinguish between the unified representatives and medical representatives involved in the case. Come. Similar to the situation of using national patent search data, corporate industrial and commercial information query data, organizational code query data, referee document query data, etc. There is no restriction on the sample of procuratorial big data, but there are indeed requirements for whether prosecutors can find information that reflects the characteristics of the case or regularity. In the era of big data today, the lack of prosecutors is not procuratorial big data, but a pair of technology perspective mirrors that understand big data. The left -looking mirror is used to find a sufficient amount of professional data or ordinary data, and the right -view mirror is used to insight from the data to the information about the relevant cases.

How to advance big data inspection?

The advancement of big data procuratorials should rely on coordinating and intertwined with scientific and technological innovation and institutional innovation to jointly form a overall operating structure.

At the level of scientific and technological innovation, when procuratorial organs organize various procuratorial big data algorithm models, the supervision of cases shall be focused. Compared with case supervision, the supervision of cases is a high -level state that cannot be split and has a significant improvement. The technique of the procuratorial organs to carry out case supervision is that in response to the common problems found in the case, the accurate and efficient supervision that promoted the solution of such problems, and finally achieved the social governance of "handling a case, supervising a batch, and governance". efficacy. In March 2022, the case No. 146 in the 36th batch of guiding cases issued by the Supreme People's Procuratorate was a good example. In the case, the traffic management department made a decision to revoke his driver's license because of the counterpart Lu Mou drunk drunk driving. Lu Mou filed an administrative lawsuit on the grounds that his behavior had been criminally judged by the court and the administrative punishment of the traffic management department. The court of the second instance judged the traffic management department. The traffic management department did not accept it and applied for the procuratorial organs to carry out legal supervision. The procuratorial organs are not handling cases, but extending the case. By analyzing the big data of the judicial data, the procuratorial organs found that the case was not a simple case together, but reflected the category of understanding and applicable to the law of the law and the application of administrative law enforcement and judicial referee. The procuratorial organs took the opportunity to protest the case, strengthened communication with the public security, courts and other departments, and applied for the revocation of administrative cases of motor vehicle driving licenses to promote the unity of judicial referee. This is a one -time solution for the incompleteness of traffic management law enforcement in drunk driving cases. The guidance significance of the case lies in "solving common issues such as inconsistency and uniform standards in law enforcement judicial cases, promoting unified law enforcement judicial standards, and correcting laws."

This kind of prosecution supervision case with policy formation is a typical case supervision. The success of the case lies in the use of big data in the administrative procuratorial supervision to form a mechanism of "discovery of case clue characteristics -potential common factor research and judgment -the construction of the case of category supervision plan". During the handling of the case, big data was first displayed. From the perspective of improving efficiency, the experience of the case can be quickly copied by developing the algorithm model of the "procuratorial big data -category supervision".

"Data collision model pointing to class cases" is the most important tool for the moment. It refers to a collision comparison of two or more datasets through special computer software, and in -depth analysis of heavy and cross -data data, and then identifies cases that can be included in the scope of supervision. Taking big data procuratorials facing the field of road traffic management as an example, my country's local procuratorate will collide with the relevant case data of the police in the public security organs on the Public Security Organ Police Comprehensive Platform. The data of the drunk driving case of the inspection agency collide with the relevant case data that was transferred to the procuratorate, or the data of the drunk driving case transferred to the prosecutor's prosecution was collided with the relevant case data transferred to the court trial. Supervision.

"Data mining model pointing to class cases" is the second largest tool. It refers to the hidden abnormal information from the algorithm from the big data, and then identifies the class case that can be included in the scope of supervision. Taking the case of vehicle insurance fraud as an example, the procuratorial organs of Shaoxing City in Zhejiang Province have used the same plaintiff or related plaintiff's complaints on traffic accident compensation as the supervision point to retrieve the legal documents of relevant local cases. It was found that the traffic accident liability disputes under the name of multiple employees in the auto repair factory were abnormally incident, and it showed a single sign of the litigation agent or appraisal of relevant cases. After that, the procuratorial organs analyzed the information flow of personnel, social security, and funds, and dug a "criminal interest chain" for defrauding insurance compensation, and then transferred this type of clue to the public security organs for investigation. If the method of finding the chain of crime interests in the case is abstracting, the data mining model of the image is obtained, and the case is directed to the class. "Data portrait of the case" is the third largest tool. The natural advantage of big data is that it can be used for portraits. The procuratorial organs can carry out data portraits of case handling groups, cases, and specific cases. As far as any type of case is concerned, a general normal distribution or law can be formed after data portraits. Taking this as a measure, if the handling of some cases is not in line with the condition or law of the normal distribution, it is necessary to transfer to the prosecutor to manually judge whether there is a tendency signs, and then decide whether to use it as a clue of the case supervision. At present, some local procuratorial organs conduct case supervision of case -type cases based on false lawsuits, civil lending, traffic accident compensation, and marriage property division. One basic premise is to set up a special data portrait model based on the characteristics of high frequency of errors in such cases. Essence

The identification of the above -mentioned procuratorial big data models of the abnormal points of cases is characterized by the automatic judgment of the "case" factors and artificial judgment of the "clue of the case". These are two tasks and two steps, indispensable.

At present, various procuratorial big data models have not reached the level of true intelligence. According to the "Opinions of the Central Committee of the Communist Party of China on Strengthening the Legal Supervision of Procuratorate in the New Era" instructions, the "use of big data, blockchain and other technologies such as big data, blockchain and other technologies" instructions, procuratorial organs can also strengthen these two steps through blockchain technology and other steps. The integration of the two tasks has prompted a case -based case supervision platform based on related models to further move towards intelligent contract supervision. Simply put, the procuratorial organs can preset various automatic execution smart contract solutions in the big data legal supervision platform or software developed by procuratorates, which not only reduces the pressure of the early prosecutor's participation, but also promotes the later procuratorial organs to collaborate with other social governance departments. Governance.

At the level of institutional innovation, my country should start the establishment of rules and regulations for supervision of procuratorial organs, and actively resolve various major legal obstacles that have or may occur. For example, big data procuratorials often have the colors of prior supervision, which shows that the introduction of big data in administrative procuratorial supervision should be alert to cross -border. The "Decision of the Central Committee of the Communist Party of China on Several major issues of the Governance of the country in accordance with the law" passed by the Fourth Plenary Session of the 18th Central Committee of the Communist Party of China. . "Here are clear the prerequisite for administrative procuratorial supervision. So, how can the procuratorial organs grasp and implement the setting of "discovery in the performance of their duties" in big data procuratorials, so that the law is evident and reasonable? These all need to be clear from the national rules and regulations.

But it is not to say that big data procuratorial can only be adhered to rules. At present, the "supervision of cases" is to achieve traceability governance, while the focus of the two is different: the former focuses on supervision, and the latter focuses on governance; The procuratorial proposal to trace the source governance, urge other social governance liability organs, social and autonomous associations and other organizations to block loopholes, and when the conditions are mature, the procuratorial organs are also involved in the coordination. To resolve potential risks from the source ... so many differences, there is a greater room for pioneering in the extension of big data procuratorial to traceability governance. In this sense, political and legal scholars are not bystanders and outsiders of big data procuratorial exploration.

Taking the category as the starting point, the transformation of the concept, the source of the data, the development of the model, the model research and development, and the system follow the system. (Prosecutor's Daily)

- END -