Personalized business data analysis empowers ＂four major prosecutors＂
Hankonglin
Wu Zhen
□A qualified data analysis report is not only a report card, but also a task list. It can objectively describe the business situation, highlight business points and focus on business problems, and set a good target and point out the direction for subsequent analysis and judgment.
□ Specific to the "Four Procuratorates", the different work promotion concepts of "being excellent, being strong, being practical and doing well" make the "Four Procuratorates" have different focus points at this stage, and corresponding personalized data analysis methods can be sorted out. When analyzing the data, according to the characteristics of different businesses and the key points of analysis, we can sum up the rules more pertinently, find out the power point, and effectively help the "four major procuratorates" to develop in an all-round way.
In recent years, under the guidance of the Supreme People’s Procuratorate, procuratorial organs at all levels across the country have gradually established a working mechanism for business data analysis, judgment and consultation. The basis of judgment consultation is business data analysis. Generally, data analysis is carried out first, and then judgment consultation is conducted. A qualified data analysis report is not only a report card, but also a task list. It can objectively describe the business situation, highlight business points and focus on business problems, and set a good target and point out the direction for subsequent analysis and judgment.
From the functional point of view, data analysis in the management of procuratorial business avoids the problem of "one leaf can’t see Mount Tai" caused by focusing too much on individual cases. Using the advantages of data perspective relative to macro and meso, we can realize in-depth analysis from "case to case, part to whole, phenomenon to essence", which can not only capture the law of procuratorial business from the data level, but also find out the handling problems and supervision clues of cases or prosecutors through the monitoring of data deviation and abnormality, and serve the specific handling of cases at the micro level. High-quality data analysis can broaden the vision of traditional procuratorial handling cases, improve the level of discovery and supervision of typed problems, create new procuratorial knowledge, and then transform it into business kinetic energy. As long as there is sufficient data support and the analysis method is quite flexible, it can respond to different analysis needs. By changing the analysis perspective and enriching the analysis methods, we can create a data analysis report that conforms to the business rules and characteristics.
Specific to the "Four Procuratorates", the different work promotion concepts of "being excellent, being strong, being practical and doing well" make the "Four Procuratorates" have different focus points at this stage, and corresponding personalized data analysis methods can be sorted out. When analyzing the data, we should sum up the rules more pertinently according to different business characteristics and analysis focuses, find out the power point, and effectively help the "four major procuratorates" to develop in an all-round way.
From the perspective of business data analysis of criminal prosecution —— Case handling quality and judicial efficiency
(A) the business characteristics and analysis focus of criminal prosecution
Criminal prosecution, as the most complex business process and the largest amount of data, is the most important part of business data analysis, and the correlation between its multiple business processes provides sufficient scenarios for business data analysis. Numerous data in criminal prosecution service is a double-edged sword in analysis. Too many analysis points mean that the time cost of analysis will be doubled and it is difficult to focus. Therefore, it is necessary to find a relatively centralized analysis model, and "doing excellent criminal prosecution" can be a breakthrough path.
In the new pattern of "Four Procuratorates", to be superior is not to pursue a larger scale, but to strive to improve the quality of handling cases and the effectiveness of supervision. Its core is "fairness", which means "trying to make the people feel fair and just in every judicial case". Fairness is the standard of justice and the direction of every case. How to achieve fairness, Han Feizi said that "the balance is balanced". In the modern rule of law, the balance of the rule of law lies not only in legislation, but also in judicial equity. In criminal justice, the difficulty of equity lies in the infinite pursuit of justice essence and the balance between limited judicial resources. "Quality and efficiency" have always been the two goals pursued by judicial value. Although they overlap and even conflict, if we can find the balance between them in individual cases or the whole judicial level, it will undoubtedly give better consideration to fairness and justice. Therefore, the focus of criminal prosecution data analysis is "fairness and justice", and the specific foothold should return to "quality and efficiency".
(2) Analysis methods of business data of criminal prosecution.
1. Cluster analysis is a common method to identify the degree of judicial balance. scatter diagram and Pareto diagram can find out the Pareto optimization of case handling quality (based on the improvement under "Pareto optimization", at least one person becomes better without making anyone’s situation worse. Specifically, it mainly refers to the space to improve case handling quality with the same judicial input). Criminal prosecution should grasp the two goals of quality and efficiency, and data analysis should also take into account two aspects. The horizontal analysis of data can better prompt the deviation of judicial fairness, realize the determination of punishment of "the same case and the same sentence" and improve the quality of cases; On the other hand, the addition of longitudinal research can reveal the occurrence law and development trend of criminal cases and predict the urgency of the demand for criminal prosecution resources in the future; At the same time, it measures the efficiency of handling criminal cases, and then provides data support for adjusting and optimizing the efficiency of handling cases.
When evaluating the quality of criminal prosecution, data analysis can pay attention to the balance of judicial scale, and various cluster analysis is an effective method to measure judicial balance. For example, whether there is a clustering difference between prosecution and relative non-prosecution through charges and circumstances (the analysis process of grouping the collection of objects into multiple classes composed of similar objects, and analyzing the differences of their grouping), and whether there is a linear clustering between the two by the amount (circumstances) and the distribution of court sentencing after prosecution, and whether there is a big difference in the distribution of sentencing in adjacent areas. If the overall judicial scale of a crime is too high after controlling variables, it shows that the judicial arbitrariness of this crime is too strong, and it is necessary to strengthen special investigation and professional guidance.
In examining the efficiency of handling cases, data analysis can objectively measure the efficiency threshold of criminal prosecution, including "optimization of investigation efficiency and supervision game", "control of case handling cycle and judicial cost", "intensity of handling cases and allocation of human resources", etc. The utility curve obtained by two-dimensional scatter diagram or Pareto diagram can identify whether there is room for optimization of efficiency, and the scatter diagram between different subjects can know the efficiency level and rising space of handling cases in our hospital. Through analysis, this paper points out the path of pursuing Pareto optimization in handling cases, so as to improve the efficiency of criminal prosecution as much as possible.
2. In the stage analysis, we should pay attention to the abnormal fluctuation of data, and the data with large deviation should be the focus of subsequent analysis. In the stage analysis (half a year or the whole year), we should pay attention to the abnormal fluctuation of the longitudinal data in the links of arrest, prosecution and judgment, and analyze the reasons for the change if the absolute value and amplitude of data rise and fall have changed greatly in the same period last year. If the adjustment of judicial standards, criminal policies and other factors is excluded, there is a great possibility of vertical imbalance in law enforcement standards, which will affect the realization of fairness and justice in case handling. At the same time, we should also pay attention to the data differences between horizontal case-handling subjects in the same time period. Although there are great differences in the basis of cooperation between public prosecutors and prosecutors and the relationship between public prosecutors and prosecutors, and there are also differences in handling habits and organizational culture, the overall data of handling cases should be in a normal distribution trend, and the concentration trend should be normal. If the data deviates greatly, it should be analyzed and judged whether there are problems in handling cases. Taking the non-arrest rate as an example, when the non-arrest rate of a certain unit is too high or too low, we should pay attention to whether there are problems such as restricting public security to report arrest, different understanding of social dangers in public prosecution, or poor guidance and investigation before arrest. If the data of reconsideration review is abnormal, it is necessary to check whether there are problems such as improper grasp of the standard of arrest and prosecution, different understanding of public prosecution or inadequate interpretation of the law in the process of non-arrest and non-prosecution
Sometimes the combination analysis between multiple indicators can help to discover the nature of data anomalies more deeply. For example, there are too many cases of "two retreats and three postponements" in units with high case-to-case ratio. Combined with the types of charges, the extension of detention period, the amount of cases handled per capita, the average period of handling cases, the data of non-prosecution and the analysis of the undertaker, we can analyze whether it is the lack of evidence collection in the investigation, the inefficiency of handling cases in the review and prosecution, or the fact that individual prosecutors are too strict with the evidence requirements for prosecution.
3. In the thematic analysis, the common sense of criminal prosecution can be tested by regression and cluster distribution. Thematic analysis focuses on a small topic and concept, which is essentially the depth data analysis under control variables. Since judicial activities are the applicable laws, laws, regulations and judicial interpretations, which are the "major premises" in judicial syllogism, remain unchanged, so by effectively controlling variables, most of the procuratorial case data should show a certain regression distribution or a centralized trend. For example, in the case of minor injury, a cross table is made based on whether to plead guilty and admit punishment and whether to reconcile, and the data of arrest and non-arrest, prosecution and non-prosecution in each table can be listed, so as to see the grasp of the crime by the local procuratorate and test the degree of leniency in pleading guilty and admitting punishment and the fulfillment of the criminal reconciliation policy. The same is true of sentencing. According to the principle of "crime and punishment should be adapted", after controlling the factors of suspects (criminal record, confession attitude and other sentencing circumstances), the sentencing of individual crimes should be positively related to the criminal circumstances, that is, the more serious the criminal circumstances, the heavier the sentencing. For example, in theft, a relatively concentrated exponential function can be fitted by making a scatter plot of the amount of theft (independent variable X) and the declared penalty (dependent variable Y), and most of the judgments should be gathered around the function, and the deviation degree can be calculated by the Y-axis difference between the scatter points and the function. If there is a big deviation from the function curve, we need to pay attention to the appropriateness of sentencing in this case. If it deviates greatly from the function curve, it may be that the sentence is too heavy, and vice versa. Another example is to distinguish plots such as attempted and not guilty in the visualization process of scatter plot, and use different shapes and colors to classify them.We can intuitively see the influence of various sentencing circumstances. In the process of inspection, we can find out whether there is any incompatibility between guilt and punishment through quality evaluation, and then start the legal supervision procedure. If a large number of stray cases occur, it shows that there is a realistic need for adjustment of criminal policy, and it is necessary to judge the level of fit between criminal policy and current social governance and provide "procuratorial early warning" for the modernization of social governance in a timely manner.
The Perspective of Business Data Analysis of Civil Prosecutions —— Clue breadth and supervision depth
(A) the business characteristics and analysis focus of civil procuratorial work
Compared with criminal prosecution, civil procuratorial business is in "development" in terms of business volume and supervision ability. In the whole judicial system, the court’s civil litigation is the most in terms of the number of cases or the types of cases, but civil procuratorial work is still a relatively weak sector in procuratorial business. To this end, the Supreme People’s Procuratorate has put forward the requirement of "strengthening civil procuratorial work", and "strengthening service" has become the core of data analysis of civil procuratorial work.
In fact, to be strong requires both quantity and quality: on the one hand, in the case of a huge base of civil adjudication, the civil procuratorial work without basic case handling capacity cannot be considered strong; On the other hand, if the civil prosecution only focuses on some low-quality, shallow-level content and fails to produce influential supervision brands and cases, then it will not be strong. Therefore, how to find the value of data, discover the deep-seated reasons behind the data through excavation, and promote the strengthening of supervision in quantity and quality is the realistic demand for data analysis put forward by civil procuratorial business.
(2) Analysis method of business data of civil prosecution.
1. Drilling for abnormal data can strengthen the effectiveness of civil procuratorial supervision. At present, the civil prosecution urgently needs the empowerment of business data analysis, which provides a mechanism for finding the reasons from appearance to essence. When the data fluctuates abnormally or deviates, we can find the substantive factors behind the judicial judgment of civil cases by digging and analyzing them in depth, and carry out targeted supervision, strive to treat both the symptoms and the root causes and solve the institutional problems. Through the integration of data and case handling, the chain reaction will be promoted to help the civil procuratorial business make a breakthrough. For example, after the conclusion of some criminal cases in judicial practice, the civil liability for infringement of property rights involved is not properly investigated and implemented, and the relevant mistakes are difficult to correct. The transfer of cases is one way, but through combing a class of phenomena and data analysis, the difference between the two types of data can be better found, and the supervision focus of typology and urgent needs of the masses can be identified. Another example is to analyze whether there is a relatively concentrated problem in supporting prosecution cases. If there are many supporting prosecutions of migrant workers for wages and they are concentrated in a certain industry at the same time, this data phenomenon suggests that there may be social governance problems in a certain field. Civil prosecution can give full play to its functions, do a good job in procuratorial supervision and rule of law services in combination with typical cases, and contribute procuratorial wisdom to the modernization of the national governance system and governance capacity.
2. Data analysis and construction of supervision clue model can empower civil prosecutors to handle cases. At present, the clues of civil procuratorial cases mainly depend on the application of the parties, which limits the breadth of clues of civil procuratorial business to a certain extent, and data analysis can solve this problem well. Grasping and modeling external data by means of informatization can help civil procuratorial business find typed business clues and expand the space of civil procuratorial business initiated by authority. At present, a successful example is the false litigation Clue Analysis and Judgment Platform built by Aerospace Science and Technology and Jiaxing Procuratorate. This platform uses the data of judgment documents and embeds the model established after data analysis to automatically push the clues of civil prosecution cases. In this process, basic data is important, but more important is the data model obtained after business data analysis. Model building relies on data analysis to sort out the factors with high correlation and form an effective algorithm; The optimization of the model also needs data analysis, and the effectiveness evaluation and verification are carried out for the quality and efficiency of the model automatically discovering data. Through the results of data analysis, civil procuratorial work can unearth a number of supervision cases that cannot be obtained under the traditional working methods, realize the efficient excavation of high-quality supervision clues, and serve and strengthen civil procuratorial work.
From the Perspective of Business Data Analysis of Administrative Prosecutions —— Administrative Quality and Supervision Effectiveness
(A) the business characteristics and analysis focus of administrative procuratorial work
On the one hand, administrative procuratorial supervision of the people’s court justice, on the other hand, to promote the administration of administrative organs according to law, "single-handedly entrusted two" let it pursue not only referee supervision, but also to promote the level of administration according to law. It is precisely for this reason that the Supreme People’s Procuratorate put forward the idea of "practical administrative procuratorial work", which is to transfer the effect of legal supervision from administrative litigation to the front end of administrative behavior, to make supervision effective, to truly "resolve administrative disputes substantively" in handling cases, and to realize "closing the case to settle politics".
Therefore, administrative procuratorial work is more concerned with "effect" than other businesses, on the one hand, it is the governance effect of specific cases in administrative law enforcement and administrative adjudication, on the other hand, it is the effect achieved by procuratorial organs in administrative procuratorial work. Data analysis should not only help to measure the effect, but also find the accurate path to improve the effect. It not only needs a high degree of integration of data and business knowledge, but also relies on the understanding of external fields such as administrative law enforcement to help test the accuracy of the conclusions drawn from the analysis of procuratorial business data. From this point of view, the business data analysis of administrative procuratorial work is the most difficult among the "Four Procuratorates".
(B) analysis methods of business data of administrative procuratorial work
1. The analysis of case-like data can help administrative prosecutors find efficient ways to handle cases. The analysis after the collection of case data can help to dig out the breakthrough of administrative procuratorial work that can not be found in a single case and the deviation of administrative law enforcement standardization from many aspects. First, the key points of business. Finding the commonness of administrative procuratorial business is an effective way to discover the essence of this kind of cases. For example, the type of specific administrative act sued, the sued unit or the plaintiff in administrative effective judgment supervision can help to understand the core of administrative litigation disputes, and the abnormal aggregation of data such as fields, units and courts will prompt the key points of administrative procuratorial supervision. For example, although the supervision of trial violation at a certain stage is accepted by different grass-roots procuratorates, it is aimed at the same court (the administrative court of grass-roots courts is centralized) and it is a similar problem. Case analysis can effectively sort out such situations and improve the effectiveness of supervision. Second, the key parties, the complainant, the accused unit and the specific types of administrative acts that often appear in administrative procuratorial work are actually highly correlated. Taking the supervision of effective judgment as an example, the case has an obvious "28 law", that is, a few parties repeatedly file a large number of administrative lawsuits that are not likely to win, and then go to the procuratorate for supervision after losing the case. Through the cluster analysis of case handling data, we can sort out at least two key points in handling cases-parties and administrative organs. The former helps us to handle cases and effectively promote the "substantive resolution of administrative disputes" by doing a good job of key people. The latter helps to implement the long-term mechanism of administrative procuratorial legal supervision,By sorting out the common types and objects of administrative procuratorial work, it helps to put forward procuratorial suggestions to administrative organs to standardize law enforcement, from handling cases to standardizing similar problems.
2. The introduction of external data analysis can help to further optimize the effect of administrative procuratorial supervision. At present, the standardization level of administrative activities really needs to be improved, especially in the field of non-litigation execution supervision, which varies greatly between cases. It is not so much the cheapness of administrative activities as the arbitrariness and lack of strict procedural consciousness of administrative organs in the execution of administrative non-litigation Therefore, the data analysis of executive supervision needs to be combined with external data and case handling data in administrative procuratorial work, taking the rationality and consistency of typed administrative executive supervision as the standard, which can not only ensure the optimization of supervision effect in procuratorial links, but also promote the effect of administrative activities and effectively promote the level of administrative effectiveness.
From the Perspective of Business Data Analysis of Public Interest Litigation Prosecution —— Supervision Effectiveness and Social Welfare
(A) the business characteristics and analysis focus of public interest litigation procuratorial work
The development of public interest litigation procuratorial has achieved a great increase in the number and scale of handling cases. While the amount of supervision is increasing, the improvement of supervision effectiveness and the increase of social welfare are the focus of the next stage of public interest litigation procuratorial work, and also the dynamic response of procuratorial organs to the urgent needs of the people for public welfare maintenance. Whether the procuratorial work is good or not, whether the procuratorial products provided are excellent or not, the most important thing is user evaluation, which is the people’s sense of acquisition and recognition of procuratorial products. Therefore, "doing a good job in public interest litigation" is not only concerned with the continuous growth of the number and scale of handling cases, but also how to better realize the repair of public interest losses.
As a new business, the business data analysis of public interest litigation lacks historical data for reference and comparison, and the business data every year updates our understanding and cognition of public interest litigation prosecution. Due to the lack of precipitation of business rules, it is difficult to grasp the reasonable interval of business data at present, and it is more difficult to analyze it. Data analysis should locate the current situation of business growth, focus on the horizontal research and cluster analysis of current business, and help the public interest litigation prosecution develop fully and evenly. At the same time, through data analysis, prominent and high-incidence problems are found, prompting the integration of supervision effectiveness, promoting the optimization of supervision effectiveness with the same investment of procuratorial resources, making up for the loss of public welfare to the greatest extent and realizing the increase of overall social welfare.
(B) Public interest litigation procuratorial business data analysis methods
1. Horizontal comparison can find the efficient range of public interest litigation. When comparing horizontal data, we should pay attention to the efficiency of clues, the effectiveness of supervision after the case is completed, and the situation of public welfare restoration after supervision. For example, in view of the difference in the success rate of clues between different hospitals, it can be analyzed whether the hospitals with low success rate put too much energy into the front-end clue arrangement, resulting in too many clues to take care of filing, or whether the filing standards are too high to grasp and miss the excavation of valuable clues; Another example is that the hospital with a high case completion rate is too harsh on clue selection, which leads to the artificial exclusion of clues that are difficult to investigate, or the standard for filing a case is too wide. The ratio of entering the pre-litigation procedure after filing a case can confirm the relevant judgment. At the same time, we should also pay attention to the problem of repeatedly carrying out administrative public interest litigation procuratorial work on the same matters at the same time, squeeze "water injection supervision" through data analysis, and reduce inefficient supervision simply to improve case handling data. Through horizontal comparison, we can find the reasonable scale of each link of public interest litigation cases and ensure that the limited inspection power is put into the most valuable clues. At the same time, we should pay attention to which data are the most concentrated, which cases involve a large amount, and which areas have an increasing trend in various fields, and make in-depth analysis to study how to optimize the quality and efficiency of public interest litigation.
2. External data analysis points out the direction for better public interest litigation prosecution. Through the analysis of some external data, it can help to find high-quality supervision clues of public interest litigation, so that public interest litigation supervision can be accurate and efficient. For example, through the collision of letters and visits and hotline data, it is found that the public has strongly reflected the clues of public interest litigation, and existing administrative data can also be used, such as the analysis of water quality sample data in the local area, and if the water quality of individual rivers in the overall water area is Grade IV or V, it is possible to actively understand the reasons for the abnormal data, whether there is any violation of public interest and the working space of the procuratorate. For the public interest litigation cases that have been handled, it is necessary to track the effectiveness of supervision through external data, verify whether the damaged public interests have really been repaired, and decide whether to carry out "looking back". Through data verification, we can ensure the effectiveness of public interest litigation prosecution, maximize social welfare, and make people feel better.
(The authors are respectively the director and prosecutor of the case management office of the Shanghai Municipal People’s Procuratorate.)