A case conclusion backstepping inspection method, an electronic device, a medium and a computer program product

Abstract

The application relates to the technical field of data backstepping verification, in particular to a case conclusion backstepping inspection method, an electronic device, a medium and a program product, the method comprises the following steps: obtaining a case analysis conclusion, and identifying relevant evidence contained in the case analysis conclusion, the case analysis conclusion is uploaded after being determined by relevant case handling personnel according to original case data; based on a trained AI identification model, the case analysis conclusion is analyzed to obtain an evidence range corresponding to the case analysis conclusion; based on the evidence range and the original case data, AI backstepping evidence is determined; the relevant evidence and the AI backstepping evidence are matched, and the case analysis conclusion is backstepped and evaluated based on the matching result. The application is convenient for improving the accuracy in the case processing process.

Description

Technical Field

[0001] This application relates to the field of data verification technology, and in particular to a method, electronic device, medium and program product for reverse verification of case conclusions. Background Technology

[0002] A complete case handling process typically includes case acceptance, investigation and evidence collection, evidence analysis, personnel review, and case reporting. The case report stage requires attaching the case conclusion. Throughout the case handling process, the case report needs to integrate information from all other stages to provide a comprehensive summary and analysis of the case. During the case report stage, relevant case handlers need to analyze the collected evidence and applicable laws to determine a conclusive opinion on the case, including but not limited to the preliminary judgment and the nature of the case.

[0003] During the case reporting stage, the analysis of collected evidence and applicable laws often relies on relevant personnel. However, when faced with a large amount of complex evidence, the sheer volume of evidence may lead to omissions of key information or misinterpretations, potentially affecting the accuracy of the case analysis results. Summary of the Invention

[0004] To improve the accuracy of case handling, this application provides a method, electronic device, medium, and program product for reverse verification of case conclusions.

[0005] Firstly, this application provides a method for reverse-engineering a case conclusion, employing the following technical solution: A method for reverse-engineering case conclusions, including: Obtain case analysis conclusions and identify relevant evidence contained in the case analysis conclusions. The case analysis conclusions are determined and uploaded by relevant case handling personnel based on the original case data. Based on the trained AI recognition model, the case analysis conclusions are analyzed to obtain the scope of evidence corresponding to the case analysis conclusions; Based on the aforementioned scope of evidence and the original case data, AI-inferred evidence is determined; The relevant evidence and the AI-backed evidence are matched, and the case analysis conclusions are evaluated based on the matching results.

[0006] By adopting the above technical solution, after the relevant case handling personnel determine the case analysis conclusion, instead of directly feeding back the determined case analysis conclusion, AI technology is used to reverse-engineer the case analysis to evaluate the accuracy of the case analysis conclusion. This facilitates the improvement of the accuracy of case handling results. By combining AI technology to analyze the case analysis conclusion, it is easier to discover any possible omissions of evidence in a timely manner. That is, by using AI technology to reverse-engineer and verify the case analysis conclusion, it is easier to verify the completeness and reliability of relevant evidence. This helps to ensure that the case analysis conclusion is obtained on the basis of sufficient and accurate evidence, thereby improving the accuracy of the case handling process.

[0007] In one possible implementation, matching the relevant evidence with the AI-backed evidence and performing a reverse evaluation of the case analysis conclusions based on the matching results includes: Identify the relevant evidence identifiers contained in the relevant evidence, and the AI ​​reverse evidence identifiers in the AI ​​reverse evidence; Based on the relevant evidence identifier and the relevant evidence, a relevant evidence group is determined for each relevant evidence identifier. Each relevant evidence group contains a relevant evidence identifier and the relevant content corresponding to the relevant evidence identifier. Based on the AI ​​back-inference evidence identifier and the AI ​​back-inference evidence, determine the AI ​​back-inference evidence group corresponding to each AI back-inference evidence identifier. Each AI back-inference evidence group contains an AI back-inference evidence identifier and the AI ​​back-inference content corresponding to the AI ​​back-inference evidence identifier. Based on the relevant evidence identifier and the AI-inferred evidence identifier, the relevant evidence group and the AI-inferred evidence group are matched to obtain the evidence matching value; When the evidence matching value is lower than the preset matching threshold, a review instruction is formed based on AI-backed evidence and the relevant evidence to achieve a back-evaluation of the case analysis conclusion.

[0008] By adopting the above technical solution, and by identifying the evidence identifiers contained in the relevant evidence and the AI-inferred evidence, it is convenient to refine the relevant evidence and the AI-inferred evidence into multiple evidence groups. Then, by comparing the evidence identifiers, the relevant evidence groups are matched with the corresponding AI-inferred evidence groups, which improves the directionality and accuracy of the matching process, thereby improving the accuracy of the matching results. When the evidence matching value is low, the case analysis conclusions are evaluated by automatically triggering a review instruction. This allows relevant case handlers to promptly identify and adjust the problems in the case analysis conclusions, thereby improving the efficiency of the case handling process.

[0009] In one possible implementation, the method further includes: Based on the relevant evidence and the AI-inferred evidence, determine the first distinguishing evidence corresponding to the relevant evidence and the second distinguishing evidence corresponding to the AI-inferred evidence; Obtain a case processing flow node diagram, and determine whether the evidentiary effects produced by the first distinguishing evidence and the second distinguishing evidence are consistent based on the case processing flow node diagram; If so, then based on historical case handling records, determine the first determination period corresponding to the first distinguishing evidence and the second determination period corresponding to the second distinguishing evidence; Based on the case processing flow node diagram, determine the first associated node information corresponding to the first distinguishing evidence and the second associated node information corresponding to the second distinguishing evidence. The associated node information includes the associated node type and the number of associated nodes. Based on the first determined period and the first associated node information, the first evidence score corresponding to the first distinguishing evidence is determined; Based on the second determined period and the second associated node information, the second evidence score corresponding to the second distinguishing evidence is determined; When the score of the second piece of evidence is higher than the score of the first piece of evidence, the case processing flowchart is optimized based on the second distinguishing evidence.

[0010] By adopting the above technical solution, when the relevant evidence group and the AI-inferred evidence group are found to be different, the specific differences between the relevant evidence group and the AI-inferred evidence group are further analyzed to understand the role and impact of different evidence on the case analysis conclusion. When the evidence effects of the two are consistent, the determination cycle and importance of the first and second differentiating evidence are quantitatively analyzed, and the case processing flow node diagram is optimized according to the differentiating evidence with higher evidence scores. This facilitates the optimization effect of the case processing flow node diagram, thereby improving the efficiency of determining the case analysis conclusion based on the optimized case processing flow node diagram.

[0011] In one possible implementation, determining whether the evidentiary effects of the first distinguishing evidence and the second distinguishing evidence are consistent based on the case processing flowchart includes: Obtain a case processing flow node diagram, and based on the case processing flow node diagram and the first distinguishing evidence, determine the first completed processing flow node, and based on the case processing flow node diagram and the second distinguishing evidence, determine the second completed processing flow node. Based on the first completed processing node and the original case data, a first simulation driving model is determined, and based on the first simulation driving model and the first distinguishing evidence, a first driving result is obtained; A second simulation driving model is determined based on the second completed processing node and the original case data, and a second driving result is obtained based on the second simulation driving model and the second distinguishing evidence; The first push result and the second push result are matched to determine the push result matching value. When the push result matching value is higher than the preset push matching threshold, it is determined that the evidence produced by the first distinguishing evidence and the second distinguishing evidence are consistent.

[0012] By adopting the above technical solution, the simulation-driven model facilitates the objective simulation of each processing stage in the case handling process. Based on the simulation results, it is easy to accurately compare whether the effects and influences of the first and second differentiating evidence are consistent, thereby improving the objectivity and accuracy of determining the comparison results. In addition, the consistency of the evidentiary effects is determined based on a preset driving matching threshold, rather than by subjective judgment, which helps to improve the accuracy of the judgment results.

[0013] In one possible implementation, the second distinguishing evidence optimizes the case processing flow node diagram, including: Identify the arrival route corresponding to the second distinguishing evidence in the case processing flow node diagram; Identify the case processing flow nodes contained in the arrival route, and determine the execution requirements corresponding to each case processing flow node based on each case processing node and historical execution data; The requirements to be executed for each case processing node are superimposed onto the arrival route to optimize the case processing node diagram.

[0014] By adopting the above technical solution, when optimizing the case processing flow node diagram, the stage of the second distinguishing evidence in the case processing flow can be quickly located according to the arrival route. In addition, by superimposing the corresponding pending requirements at each case processing flow node in the arrival route, it is convenient to remind relevant personnel of the potential needs corresponding to each case processing flow node, thereby improving the efficiency of relevant personnel to reach the corresponding operation stage of the second distinguishing evidence through the arrival route.

[0015] In one possible implementation, the method further includes: When it is detected that the user's current execution node is any case processing flow node in the arrival route, obtain the operation information corresponding to the user at the current execution node; Based on the operation information and the pending requirements corresponding to the current execution node, determine the prompt operation information and feed it back; After receiving the execution operation information uploaded by the user, the pending requirements corresponding to the current execution node are updated based on the execution operation information. The execution operation information is uploaded by the user after receiving the prompt operation information.

[0016] By adopting the above technical solution, the real-time monitoring of user operation information at execution nodes facilitates the clear understanding of the user's real-time progress in the case handling process. In addition, updating the pending execution requirements of the corresponding execution nodes through the execution operation information uploaded by the user further clarifies the user's completion status at the current execution node. By increasing the transparency of the operation process, it is easier to promptly identify non-compliant or inappropriate behaviors in the process of determining case analysis conclusions, thereby improving the accuracy of determining case analysis conclusions.

[0017] Secondly, this application provides an electronic device that adopts the following technical solution: An electronic device comprising: At least one processor; Memory; At least one application, wherein the at least one application is stored in memory and configured to be executed by at least one processor, the at least one application being configured to: execute the aforementioned case conclusion reverse verification method.

[0018] Thirdly, this application provides a computer-readable storage medium, which adopts the following technical solution: A computer-readable storage medium includes: a computer program that can be loaded by a processor and execute the aforementioned case conclusion reverse verification method.

[0019] Fourthly, this application provides a computer program product, which adopts the following technical solution: A computer program product includes a computer program that, when executed by a processor, implements the aforementioned method for reverse-engineering the conclusions of cases.

[0020] In summary, this application includes at least one of the following beneficial technical effects: After the relevant case handling personnel determine the case analysis conclusion, instead of directly providing feedback on the determined case analysis conclusion, they use AI technology to reverse-engineer the case analysis to evaluate the accuracy of the case analysis conclusion. This facilitates the improvement of the accuracy of the case handling results. By combining the case analysis conclusion with AI technology, it is easier to promptly discover any possible omissions of evidence. In other words, the reverse-engineering and verification of the case analysis conclusion with AI technology facilitates the verification of the completeness and reliability of relevant evidence. This ensures that the case analysis conclusion is obtained on the basis of sufficient and accurate evidence, thereby improving the accuracy of the case handling process.

[0021] When optimizing the case processing flow node diagram, the arrival route facilitates quick location of the stage of the second distinguishing evidence in the case processing flow. In addition, by overlaying the corresponding pending requirements at each case processing flow node in the arrival route, it is easy to remind relevant personnel of the potential needs corresponding to each case processing flow node, thereby improving the efficiency of relevant personnel in reaching the corresponding operation stage of the second distinguishing evidence through the arrival route. Attached Figure Description

[0022] Figure 1 This is a flowchart illustrating a method for reverse-engineering a case conclusion in an embodiment of this application; Figure 2 This is a schematic diagram of a reverse evaluation process in an embodiment of this application; Figure 3 This is a schematic diagram of an optimized case handling process node diagram in an embodiment of this application; Figure 4 This is a schematic diagram of the structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0023] The following is in conjunction with the appendix Figure 1-4 This application will be described in further detail.

[0024] After reading this specification, those skilled in the art may make modifications to this embodiment without contributing any inventive step, but such modifications are protected by patent law as long as they fall within the scope of the claims of this application.

[0025] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0026] It should be noted that, in the optional embodiments of this application, the data related to object information, when applied to specific products or technologies, requires the permission or consent of the object. Furthermore, the collection, use, and processing of this data must comply with the relevant laws, regulations, and standards of the relevant countries and regions. In other words, if the embodiments of this application involve data related to an object, it must be obtained with the object's authorization and consent, the authorization and consent of relevant departments, and in accordance with the relevant laws, regulations, and standards of the country and region. If the embodiments involve personal information, the acquisition of all personal information requires the individual's consent. If sensitive information is involved, the separate consent of the information subject is required. The embodiments also need to be implemented with the object's authorization and consent.

[0027] Specifically, this application provides a method for reverse-engineering a case conclusion, executed by an electronic device. This electronic device can be a server or a terminal device. The server can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services. The terminal device can be a smartphone, tablet, laptop, desktop computer, etc., but is not limited to these. The terminal device and the server can be directly or indirectly connected via wired or wireless communication, and this application does not impose any limitations on this.

[0028] refer to Figure 1 , Figure 1 This is a flowchart illustrating a case conclusion reverse verification method according to an embodiment of this application. The method includes steps S110-S140, wherein: Step S110: Obtain the case analysis conclusion and identify the relevant evidence contained in the case analysis conclusion. The case analysis conclusion is determined by the relevant case handling personnel based on the original case data and then uploaded.

[0029] Specifically, the original case data is used to describe the cases to be handled. This original case data may include basic case information, such as case number, case type, case name, case status, and case source; it may include party information, such as the parties' names, genders, ages, contact information, and addresses; it may include case process records, such as detailed records of each stage, including investigation, trial, and judgment; and it may include case-related information, such as information on other related cases and industry standards, to provide a broader analytical perspective for case handling. Because different cases occur under different backgrounds and environments, the original case data corresponding to different cases will differ. The specific content of the original case data is not specifically limited in this embodiment.

[0030] Relevant case handlers can use their experience and legal knowledge to perform legal reasoning on the original case data to obtain corresponding case analysis conclusions. These conclusions generally include the determination of facts, relevant evidence, social impact analysis, and legal application analysis. After determining the case analysis conclusions corresponding to the original case data, the handlers can upload them to an electronic device via a terminal device. When identifying relevant evidence in the case analysis conclusions, feature recognition algorithms can be used. The specific recognition method is not limited in this embodiment, as long as the relevant evidence contained in the case analysis conclusions can be identified.

[0031] Step S120: Based on the trained AI recognition model, analyze the case analysis conclusions to obtain the scope of evidence corresponding to the case analysis conclusions.

[0032] Specifically, the AI ​​recognition model is trained on a large amount of sample data. This sample data includes different sample cases and corresponding manual annotations. The manual annotations include the scope of evidence for the sample cases and the corresponding case analysis conclusions. The scope of evidence corresponds to the case analysis conclusions; that is, the case analysis conclusions can be obtained from the scope of evidence. The AI ​​recognition model, trained on a large amount of sample data, can identify the corresponding scope of evidence from case analysis conclusions of different types of cases. When the case analysis results are imported into the trained AI recognition model, it can automatically identify the scope of evidence corresponding to the case analysis conclusions. The scope of evidence includes, but is not limited to, relevant factual descriptions, legal clause citations, and evidence names. The training method for the AI ​​recognition model is not specifically limited in this embodiment, as long as the trained AI recognition model can obtain the scope of evidence corresponding to the factual findings in the case analysis conclusions.

[0033] Step S130: Based on the scope of evidence and the original case data, determine the evidence that AI can reverse-engineer.

[0034] Specifically, based on the scope of evidence, the original case data is filtered to obtain AI-inferred evidence corresponding to that scope. In other words, an AI recognition model determines the AI-inferred evidence corresponding to the case analysis conclusions. Relevant evidence is obtained by the case handlers based on the original case data analysis, while AI-inferred evidence is derived by using the AI ​​recognition model to infer the case analysis conclusions from those conclusions. Relevant evidence and AI-inferred evidence may be the same or different.

[0035] Step S140: Match the relevant evidence with the AI-backed evidence, and evaluate the case analysis conclusions based on the matching results.

[0036] Specifically, since the training and prediction of AI recognition models mainly rely on the sample data corresponding to the model training stage, when using a trained AI recognition model for reverse inference, it mainly depends on the logic between the data and the algorithm. However, when the relevant case handlers determine the case analysis conclusions, they mainly rely on their legal practice and experience. Therefore, the relevant evidence contained in the case analysis conclusions determined by the relevant case handlers may be different from the AI ​​reverse inference evidence derived from the AI ​​recognition model.

[0037] The relevant evidence is matched with the AI-generated evidence. A high match value indicates that the evidence used by the case handlers to determine the facts of the case is quite similar to the AI-generated evidence. In this case, the case analysis conclusion can be directly fed back. A low match value indicates that there is a significant difference between the evidence used by the case handlers to determine the facts of the case and the AI-generated evidence. In this case, the case analysis conclusion will not be uploaded directly. Instead, the AI-generated evidence and the case analysis conclusion will be fed back to the relevant evaluation personnel to prompt them to evaluate the case classification conclusion determined by the case handlers.

[0038] In this embodiment of the application, after the relevant case handlers determine the case analysis conclusion, the determined case analysis conclusion is not directly fed back. Instead, AI technology is used to reverse-engineer the case analysis to evaluate the accuracy of the case analysis conclusion, thereby improving the accuracy of the case handling results. By combining AI technology to analyze the case analysis conclusion, it is easier to discover any possible omissions of evidence in a timely manner. That is, by using AI technology to reverse-engineer and verify the case analysis conclusion, it is easier to verify the completeness and reliability of the relevant evidence, thereby ensuring that the case analysis conclusion is obtained on the basis of sufficient and accurate evidence, and thus improving the accuracy of the case handling process.

[0039] Furthermore, to facilitate timely identification of potential problems in case analysis conclusions by relevant case handlers, relevant evidence is matched with AI-generated evidence. Based on the matching results, the case analysis conclusions are evaluated using a reverse-engineering method, including steps S1401-S1405, such as... Figure 2 As shown, where: Step S1401: Identify the relevant evidence identifiers contained in the relevant evidence and the AI ​​reverse evidence identifiers in the AI ​​reverse evidence.

[0040] Specifically, different relevant evidences correspond to different relevant evidence identifiers. Similarly, different AI-inferred evidences correspond to different AI-inferred evidence identifiers. For example, the relevant evidence identifier for relevant evidence 1 can be X01, and the relevant evidence identifier for relevant evidence 2 can be X02. The AI-inferred evidence identifier for AI-inferred evidence 1 can be A01, and the AI-inferred evidence identifier for AI-inferred evidence 2 can be A02. The relevant evidence identifiers and AI-inferred evidence identifiers can be displayed as codes or symbols. The specific display format is not specifically limited in this application embodiment, as long as the corresponding relevant evidence can be uniquely identified based on the relevant evidence identifier. Similarly, as long as the corresponding AI-inferred evidence can be uniquely identified based on the AI-inferred evidence identifier, it is acceptable.

[0041] Step S1402: Based on the relevant evidence identifier and the relevant evidence, determine the relevant evidence group corresponding to each relevant evidence identifier. Each relevant evidence group contains a relevant evidence identifier and the relevant content corresponding to the relevant evidence identifier.

[0042] Specifically, relevant evidence groups are determined based on relevant evidence identifiers. That is, different pieces of relevant evidence are distinguished based on relevant evidence identifiers, and each relevant evidence group contains a relevant evidence identifier and the relevant evidence corresponding to that identifier. One relevant evidence group corresponds to one relevant evidence identifier. This grouping facilitates intuitive viewing and analysis of each piece of relevant evidence.

[0043] Step S1403: Based on the AI ​​reverse evidence identifier and the AI ​​reverse evidence, determine the AI ​​reverse evidence group corresponding to each AI reverse evidence identifier. Each AI reverse evidence group contains an AI reverse evidence identifier and the AI ​​reverse content corresponding to the AI ​​reverse evidence identifier.

[0044] Specifically, the method for determining AI-inferred evidence groups can refer to the method for determining relevant evidence groups, and will not be repeated here. Each AI-inferred evidence group corresponds to an AI-inferred evidence identifier. Since relevant evidence may differ from AI-inferred evidence, the number of relevant evidence groups and the number of AI-inferred evidence groups may also differ.

[0045] Step S1404: Based on the relevant evidence identifier and the AI-inferred evidence identifier, match the relevant evidence group and the AI-inferred evidence group to obtain the evidence matching value.

[0046] Specifically, when matching relevant evidence identifiers with AI-inferred evidence identifiers, it is necessary to first determine the AI-inferred evidence identifier corresponding to the relevant evidence identifier from multiple AI-inferred evidence identifiers. For example, the AI-inferred evidence identifier corresponding to relevant evidence identifier X01 is A01, not A02. If the relevant evidence identifier and its corresponding AI-inferred evidence identifier share the same identifier code, after determining the AI-inferred evidence identifier corresponding to the relevant evidence identifier, the evidence sets corresponding to both can be combined to obtain an evidence set. The evidence sets contained within the evidence set are then matched to obtain the set evidence matching value. When there is no AI-inferred evidence identifier corresponding to a relevant evidence identifier, the evidence set corresponding to that relevant evidence identifier only contains the relevant evidence identifier group corresponding to that relevant evidence identifier, and the set evidence matching value corresponding to that evidence set is 0. Similarly, when there is no relevant evidence identifier group corresponding to an AI-inferred evidence identifier, the set evidence matching value corresponding to that AI-inferred evidence identifier is 0.

[0047] After identifying all relevant evidence identifiers and AI-inferred evidence identifiers, calculate the average matching value of all aggregated evidence matching values, and determine the average matching value as the evidence matching value.

[0048] Step S1405: When the evidence matching value is lower than the preset matching threshold, a review instruction is formed based on AI to reverse-engineer the evidence and related evidence, so as to reverse-engineer the case analysis conclusion.

[0049] Specifically, the preset matching threshold can be 98% or 95%. The specific value of the preset matching threshold is not limited in this embodiment and can be set by relevant technical personnel according to actual needs. Once the evidence matching value is lower than the preset matching threshold, it indicates that there is a significant difference between the AI-inferred evidence and the relevant evidence. At this time, there may be evidence missing or omission. Therefore, it is necessary to generate a review instruction to remind relevant personnel to re-evaluate or review the case analysis conclusions to ensure the accuracy of the case analysis conclusions.

[0050] Furthermore, to improve efficiency in determining case analysis conclusions based on the optimized case processing flowchart, the method provided in this application embodiment further includes steps S1-S7, such as... Figure 3 As shown, where: Step S1: Based on the relevant evidence and the AI-inferred evidence, determine the first distinguishing evidence corresponding to the relevant evidence and the second distinguishing evidence corresponding to the AI-inferred evidence.

[0051] Specifically, the first distinguishing evidence is the relevant evidence that differentiates from the AI-inferred evidence. Correspondingly, the second distinguishing evidence is the AI-inferred evidence that differentiates from the relevant evidence. For example, if the relevant evidence includes a, b, and c, and the AI-inferred evidence includes a, b, d, and e, then the first distinguishing evidence can be c, and the second distinguishing evidence can be d and e. To determine the first or second distinguishing evidence, the relevant evidence can first be defined as a set of relevant evidence, and the AI-inferred evidence as a set of AI-inferred evidence. Since sets are data structures that do not allow duplicate elements, converting the relevant evidence and AI-inferred evidence into sets improves efficiency in the evidence comparison process. Then, by using the difference operation of the sets, evidence that exists in the set of relevant evidence but not in the set of AI-inferred evidence is identified and determined as the first distinguishing evidence. Similarly, evidence that exists in the set of AI-inferred evidence but not in the set of relevant evidence is identified and determined as the second distinguishing evidence.

[0052] Step S2: Obtain the case processing flow node diagram, and determine whether the evidentiary effects produced by the first distinguishing evidence and the second distinguishing evidence are consistent based on the case processing flow node diagram.

[0053] Specifically, the case handling flowchart serves as a mind map for relevant case handlers during the case processing. Two case handling branches can be created within the flowchart, based on the first and second differentiating evidence respectively. By comparing the key transition points in these two branches and analyzing the similarities and differences between them, it becomes easier to determine whether the evidentiary effects of the first and second differentiating evidence are consistent. To improve the accuracy of the judgment, determining whether the evidentiary effects of the first and second differentiating evidence are consistent based on the case handling flowchart can specifically include: Obtain a case processing flow node diagram, and based on the case processing flow node diagram and the first distinguishing evidence, determine the first completed processing flow node; based on the case processing flow node diagram and the second distinguishing evidence, determine the second completed processing flow node; based on the first completed processing flow node and the original case data, determine the first simulated driving model, and based on the first simulated driving model and the first distinguishing evidence, obtain the first driving result; based on the second completed processing flow node and the original case data, determine the second simulated driving model, and based on the second simulated driving model and the second distinguishing evidence, obtain the second driving result; match the first driving result and the second driving result to determine the driving result matching value; when the driving result matching value is higher than the preset driving matching threshold, it is determined that the evidentiary effects produced by the first distinguishing evidence and the second distinguishing evidence are consistent.

[0054] Specifically, besides comparing the evidentiary effects of the first and second differentiating evidence through two case processing branches, a simulation-driven model can also be used for judgment. This involves determining the execution logic and requirements corresponding to each processing node in the case processing flow diagram. The execution logic and requirements of each processing node are then matched with the first differentiating evidence to determine the first pending processing node. This first pending processing node requires the input of the corresponding first differentiating evidence, and there may be multiple such nodes. The first completed processing node is any processing node in the case processing flow diagram other than the first pending processing node. The method for determining the second completed processing node is similar to the method for determining the first completed processing node, and will not be elaborated upon here.

[0055] After identifying the first completed processing node, a corresponding first simulated driving model can be determined based on the first completed processing node. The process of determining the first simulated driving model may include: Based on the original case data, a corresponding basic simulation driving model is determined. This basic simulation driving model is adapted to the original case data and can simulate the case derivation process corresponding to the original case data. Different original case data correspond to different basic simulation driving models. After determining the basic simulation driving model, the driving status of the basic simulation driving model can be updated based on the first completed processing flow node to obtain the first basic simulation driving model. The simulation processing flow nodes corresponding to the first completed processing flow node in the first basic simulation driving model are also in a completed state. Finally, the first driving result is obtained by driving the optimized first basic simulation driving model with the first distinguishing evidence. Based on the above method, the second driving result is obtained by driving the optimized second basic simulation driving model with the second distinguishing evidence. The method of determining and optimizing the second basic simulation driving model is not specifically limited in this embodiment; it can be referred to the above method of determining and optimizing the first basic simulation driving model, and will not be elaborated here.

[0056] By comparing the matching values ​​of the first and second initiating results, it is easier to assess whether the evidentiary effects of the first and second differentiating evidence are consistent. Since the first initiating result represents the evidentiary effect corresponding to the first differentiating evidence, and the second initiating result represents the evidentiary effect corresponding to the second differentiating evidence, determining the matching values ​​of the initiating results allows for a quantitative analysis of the evidentiary effects of the first and second differentiating evidence, thus improving the accuracy of determining whether the evidentiary effects of the first and second differentiating evidence are consistent.

[0057] Step S3: If so, then based on the historical case handling records, determine the first determination period corresponding to the first distinguishing evidence and the second determination period corresponding to the second distinguishing evidence.

[0058] Specifically, when the first and second differentiating evidence produce the same evidentiary effect, they are considered parallel evidence and do not affect the accuracy of the case analysis conclusion. In this case, the priority of the first and second differentiating evidence can be determined by analyzing them, so as to improve the speed at which relevant case handlers determine the case analysis conclusion without affecting the accuracy of the case analysis conclusion.

[0059] Historical case records contain the determination period required to obtain or confirm any distinguishing evidence. These records facilitate the determination of the first determination period for the first distinguishing piece of evidence and the second determination period for the second distinguishing piece of evidence. Because different types of distinguishing evidence and their availability differ, the determination periods for different types of distinguishing evidence vary. A shorter determination period indicates less difficulty in determining the corresponding distinguishing evidence.

[0060] Step S4: Based on the processing flow node diagram, determine the first associated node information corresponding to the first distinguishing evidence and the second associated node information corresponding to the second distinguishing evidence. The associated node information includes the associated node type and the number of associated nodes.

[0061] Specifically, regarding the first distinguishing evidence, when determining the first related node information corresponding to the first distinguishing evidence, the distinguishing process node corresponding to the first distinguishing evidence can be identified from the case handling process nodes. The upstream and downstream nodes of the distinguishing process node are then identified as related nodes of the distinguishing process node, and the number of related nodes is determined based on this. Related node types include active related types and passive related types. The related node type corresponding to the upstream node of the distinguishing process node can be identified as the active related type, and the related node type corresponding to the downstream node can be identified as the passive related type.

[0062] Based on the above method, the first associated node information corresponding to the first distinguishing evidence and the second associated node information corresponding to the second distinguishing evidence can be determined.

[0063] Step S5: Based on the first determination period and the first associated node information, determine the first evidence score corresponding to the first distinguishing evidence.

[0064] Step S6: Based on the second determination period and the second associated node information, determine the second evidence score corresponding to the second distinguishing evidence.

[0065] Specifically, regarding the first distinguishing evidence, when determining the corresponding first evidence score based on the first determined period and the first associated node information, the first period score corresponding to the first determined period can be determined firstly based on the first determined period and the period mapping relationship. The longer the first determined period, the lower the corresponding first period score. The period mapping relationship is the correspondence between the first determined period and the first period score. The specific content of the period mapping relationship is not specifically limited in this application embodiment. Based on the period mapping relationship, the first period score corresponding to any first determined period can be determined. The method for determining the second period score corresponding to the second determined period can refer to the above-mentioned disclosure and will not be repeated here.

[0066] Regarding the first distinguishing evidence, after determining the first period score corresponding to the first defined period, the proportion of each association type can be determined based on the types and number of association nodes in the first association node information. Then, based on the first association node information and the association mapping relationship, the first association score corresponding to the first association node information is determined. The association mapping relationship includes the first association scores corresponding to different association types and different numbers of association nodes. The higher the proportion of active and passive association types, and the more association nodes there are, the higher the importance, and the higher the corresponding first association score. The specific content of the association mapping relationship is not specifically limited in this application embodiment; the first association score corresponding to any first association node information can be derived based on the association mapping relationship. The method for determining the second association score corresponding to the second defined period can refer to the above-mentioned disclosure and will not be repeated here.

[0067] The first evidence score corresponding to the first distinguishing evidence is the sum of the first period score and the first correlation score; the second evidence score corresponding to the second distinguishing evidence is the sum of the second period score and the second correlation score.

[0068] Step S7: When the score of the second piece of evidence is higher than the score of the first piece of evidence, optimize the case processing flow node diagram based on the second distinguishing evidence.

[0069] Specifically, when the score of the second piece of evidence is higher than that of the first piece of evidence, it indicates that the determination period required to determine the second piece of evidence is shorter, and the second piece of evidence is of higher importance. Therefore, when the first piece of evidence and the second piece of evidence are in parallel, relevant case handlers can be guided to prioritize the case handling branch corresponding to the second piece of evidence.

[0070] When optimizing the case processing flowchart based on second distinguishing evidence, the case processing branches corresponding to the second distinguishing evidence can be marked to remind relevant case handlers. Furthermore, to improve the case processing efficiency of relevant case handlers, optimizing the case processing flowchart based on second distinguishing evidence can specifically include: Identify the arrival route corresponding to the second distinguishing evidence in the case processing flow node diagram; identify the case processing flow nodes contained in the arrival route, and determine the execution requirements corresponding to each case processing flow node based on each case processing flow node and historical execution data; overlay the execution requirements corresponding to each case processing flow node onto the arrival route to complete the optimization of the case processing flow node diagram.

[0071] Specifically, the arrival route corresponding to the second distinguishing evidence can be the case processing branch corresponding to the second distinguishing evidence, that is, the route where node flow can only occur after the second distinguishing evidence is obtained or determined. This can be achieved by determining the second completed processing node corresponding to the second distinguishing evidence, and then determining the arrival route corresponding to the second distinguishing evidence based on the associated node information corresponding to each second completed processing node. The method of determining the arrival route is not specifically limited in this application embodiment, as long as the arrival route corresponding to the second distinguishing evidence can be determined.

[0072] Once the arrival route is determined, all case processing process nodes included in the arrival route can be identified. Historical execution data contains the pending requirements corresponding to the case handling personnel in different case processing process nodes during historical periods. The pending requirements are different for different case processing process nodes. The pending requirements include, but are not limited to, pending operations and legal knowledge to be learned. The content of the pending requirements is not specifically limited in this application embodiment.

[0073] When overlaying pending requirements onto the arrival route, the area to be overlaid for each requirement can be identified. AI technology is then used to overlay these requirements. The resulting case processing flow node diagram shows that each case processing node on the arrival route has its corresponding pending requirement overlaid. By overlaying the corresponding pending requirements at each case processing node on the arrival route, relevant personnel are alerted to the potential needs at each node, thereby improving the efficiency of reaching the second distinguishing evidence-related operational stage via the arrival route.

[0074] The display format of the area to be overlaid in the arrival route for the requirements to be executed can be different. It can be a list or a chart. The specific display format is not specifically limited in this embodiment.

[0075] The method provided in this application embodiment further includes: when it is detected that the user's current execution node is any case processing process node in the arrival route, obtaining the operation information corresponding to the user at the current execution node; determining the prompt operation information according to the operation information and the pending execution requirements corresponding to the current execution node, and feeding back the prompt operation information; after receiving the execution operation information uploaded by the user, updating the pending execution requirements corresponding to the current execution node based on the execution operation information, wherein the execution operation information is uploaded by the user after receiving the prompt operation information.

[0076] Specifically, when it is detected that a user is handling a case based on the optimized case handling process node diagram, the execution node corresponding to the user at the current moment is obtained, that is, the current execution node. The node identifier corresponding to the current execution node is matched with the node identifiers corresponding to all case handling process nodes in the arrival route to determine whether the current execution node is any case handling process node in the arrival route. The user can be the relevant case handling personnel who determine the case analysis conclusion based on the optimized case handling process node diagram.

[0077] When the current execution node is determined to be any case processing node in the arrival route, the user's operation information corresponding to the current execution node is obtained according to the operation log. The operation information includes, but is limited to, the content uploaded or downloaded by the user at the current execution node. The operation information is compared with the pending requirements corresponding to the current execution node to determine the user's non-operation information at the current execution node. Based on the non-operation information, prompt operation information is determined. The prompt operation information can be the same as the non-operation information, or the non-operation information can be extracted by keywords and the prompt operation information can be determined based on the extracted keywords. The specific form of the prompt operation information is not specifically limited in this embodiment of the application, as long as it can prompt the user. The generated prompt operation information can be fed back to the user's corresponding terminal device in the form of SMS or pop-up window.

[0078] After receiving the feedback prompt, the received uploaded content is identified as the user-uploaded execution operation information. This information may or may not correspond to the prompt. That is, the uploaded information may or may not conform to the upload standards or format specified in the prompt. Since the prompt includes the upload standards or format, it's necessary to compare the received execution operation information with the prompt to determine the standard execution operation information that meets the required upload standards or format. Once the standard execution operation information is determined, any unexecuted information corresponding to it is removed from the corresponding pending requirements, or a completed mark is added to the unexecuted information. This allows users to visually determine the completed and unexecuted information by viewing the display of the pending requirements, thus clarifying the user's execution completion rate at the current execution node. If there is no unexecuted information in the pending requirements for the current execution node, the user's execution completion rate at the current execution node is determined to be 100%. By increasing the transparency of the operational process, it is easier to promptly identify non-compliant or improper behaviors in the process of determining case analysis conclusions, thereby improving the accuracy of determining case analysis conclusions.

[0079] This application provides an electronic device, such as... Figure 4 As shown, Figure 4 The illustrated electronic device 400 includes a processor 401 and a memory 403. The processor 401 and the memory 403 are connected, for example, via a bus 402. Optionally, the electronic device 400 may also include a transceiver 404. It should be noted that in practical applications, the transceiver 404 is not limited to one type, and the structure of this electronic device 400 does not constitute a limitation on the embodiments of this application.

[0080] Processor 401 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 401 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0081] Bus 402 may include a pathway for transmitting information between the aforementioned components. Bus 402 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. Bus 402 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 4 The symbol is represented by only one line, but this does not mean that there is only one bus or one type of bus.

[0082] The memory 403 may be a ROM (Read Only Memory) or other type of static storage device capable of storing static information and instructions, RAM (Random Access Memory) or other type of dynamic storage device capable of storing information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), CD-ROM (Compact Disc Read Only Memory) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto.

[0083] The memory 403 is used to store application code that executes the solution of this application, and its execution is controlled by the processor 401. The processor 401 is used to execute the application code stored in the memory 403 to implement the content shown in the foregoing method embodiments.

[0084] Electronic devices include, but are not limited to: mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), and in-vehicle terminals (such as in-vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. Servers can also be included. Figure 4 The electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0085] This application provides a computer-readable storage medium storing a computer program that, when run on a computer, enables the computer to execute the corresponding content in the aforementioned method embodiments.

[0086] This application provides a computer program product including a computer program that, when executed by a processor, implements the methods described in any of the above embodiments.

[0087] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0088] The above description is only a partial embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for reverse-engineering case conclusions, characterized in that, include: Obtain case analysis conclusions and identify relevant evidence contained in the case analysis conclusions. The case analysis conclusions are determined and uploaded by relevant case handling personnel based on the original case data. Based on the trained AI recognition model, the case analysis conclusions are analyzed to obtain the scope of evidence corresponding to the case analysis conclusions; Based on the aforementioned scope of evidence and the original case data, AI-inferred evidence is determined; The relevant evidence and the AI-backed evidence are matched, and the case analysis conclusions are evaluated based on the matching results. The process of matching the relevant evidence with the AI-generated evidence, and then evaluating the case analysis conclusions based on the matching results, includes: Identify the relevant evidence identifiers contained in the relevant evidence, and the AI ​​reverse evidence identifiers in the AI ​​reverse evidence; Based on the relevant evidence identifier and the relevant evidence, a relevant evidence group is determined for each relevant evidence identifier. Each relevant evidence group contains a relevant evidence identifier and the relevant content corresponding to the relevant evidence identifier. Based on the AI ​​back-inference evidence identifier and the AI ​​back-inference evidence, determine the AI ​​back-inference evidence group corresponding to each AI back-inference evidence identifier. Each AI back-inference evidence group contains an AI back-inference evidence identifier and the AI ​​back-inference content corresponding to the AI ​​back-inference evidence identifier. Based on the relevant evidence identifier and the AI-inferred evidence identifier, the relevant evidence group and the AI-inferred evidence group are matched to obtain the evidence matching value; When the evidence matching value is lower than the preset matching threshold, a review instruction is formed based on AI-backed evidence and the relevant evidence to achieve a back-evaluation of the case analysis conclusion.

2. The method for reverse-engineering a case conclusion according to claim 1, characterized in that, Also includes: Based on the relevant evidence and the AI-inferred evidence, determine the first distinguishing evidence corresponding to the relevant evidence and the second distinguishing evidence corresponding to the AI-inferred evidence; Obtain a case processing flow node diagram, and determine whether the evidentiary effects of the first distinguishing evidence and the second distinguishing evidence are consistent based on the case processing flow node diagram; If so, then based on historical case handling records, determine the first determination period corresponding to the first distinguishing evidence and the second determination period corresponding to the second distinguishing evidence; Based on the case processing flow node diagram, determine the first associated node information corresponding to the first distinguishing evidence and the second associated node information corresponding to the second distinguishing evidence. The associated node information includes the associated node type and the number of associated nodes. Based on the first determined period and the first associated node information, the first evidence score corresponding to the first distinguishing evidence is determined; Based on the second determined period and the second associated node information, the second evidence score corresponding to the second distinguishing evidence is determined; When the score of the second piece of evidence is higher than the score of the first piece of evidence, the case processing flowchart is optimized based on the second distinguishing evidence.

3. The method for reverse-engineering a case conclusion according to claim 2, characterized in that, The step of determining whether the evidentiary effects of the first distinguishing evidence and the second distinguishing evidence are consistent based on the case handling flowchart includes: Obtain a case processing flow node diagram, and based on the case processing flow node diagram and the first distinguishing evidence, determine the first completed processing flow node, and based on the case processing flow node diagram and the second distinguishing evidence, determine the second completed processing flow node. Based on the first completed processing node and the original case data, a first simulation driving model is determined, and based on the first simulation driving model and the first distinguishing evidence, a first driving result is obtained; A second simulation driving model is determined based on the second completed processing node and the original case data, and a second driving result is obtained based on the second simulation driving model and the second distinguishing evidence; The first push result and the second push result are matched to determine the push result matching value. When the push result matching value is higher than the preset push matching threshold, it is determined that the evidence produced by the first distinguishing evidence and the second distinguishing evidence are consistent.

4. The method for reverse-engineering a case conclusion according to claim 2, characterized in that, The second distinguishing evidence optimizes the case processing flowchart, including: Identify the arrival route corresponding to the second distinguishing evidence in the case processing flow node diagram; Identify the case processing flow nodes contained in the arrival route, and determine the execution requirements corresponding to each case processing flow node based on each case processing node and historical execution data; The requirements to be executed corresponding to each case processing node are superimposed onto the arrival route to optimize the case processing node diagram.

5. The method for reverse-engineering a case conclusion according to claim 4, characterized in that, Also includes: When it is detected that the user's current execution node is any case processing flow node in the arrival route, obtain the operation information corresponding to the user at the current execution node; Based on the operation information and the pending requirements corresponding to the current execution node, determine the prompt operation information and feed it back; After receiving the execution operation information uploaded by the user, the pending requirements corresponding to the current execution node are updated based on the execution operation information. The execution operation information is uploaded by the user after receiving the prompt operation information.

6. An electronic device, characterized in that, The electronic device includes: At least one processor; Memory; At least one application, wherein the at least one application is stored in memory and configured to be executed by at least one processor, the at least one application being configured to: perform a case conclusion reverse verification method according to any one of claims 1-5.

7. A computer-readable storage medium, characterized in that, include: The computer program is stored and can be loaded by a processor and executed as described in any one of claims 1-5.

8. A computer program product, characterized in that, It includes a computer program that, when executed by a processor, implements the steps of a case conclusion reverse verification method according to any one of claims 1-5.

Images