Archive opening auditing method and system and storage medium

Abstract

The invention provides a file open auditing method and system and a storage medium, and the method comprises the steps: obtaining an auditing work plan, training a preset auditing model based on the auditing work plan, and obtaining a localized auditing model; receiving auditing data, and determining target personnel participating in auditing based on the auditing data to perform security screening; performing quality processing on the auditing data to obtain digital auditing data; based on the digital auditing data and the localized auditing model, obtaining a preliminary auditing opinion; generating a reexamination instruction, and receiving a reexamination result; compiling a data set based on the reexamination result and the digital reexamination data, training a localized reexamination model by using the data set to obtain an updated reexamination model, judging whether the updated reexamination model meets the reexamination accuracy or not, if so, determining the updated reexamination model as the localized reexamination model, and outputting the reexamination result; and if not, training a preset auditing model based on the auditing work plan to obtain a localized auditing model. The method can improve the situation that a large amount of manpower resources are consumed for archive auditing work.

Description

Technical Field

[0001] The present application relates to the technical field of archive review, and in particular to an archive open review method, system and storage medium. Background Art

[0002] Archives opening review refers to the activities of national archives, archives forming or transferring units, following the set work procedures and based on determined classification standards, to review and verify the content and form of the archives in their collections volume by volume, item by item, and page by page, so as to scientifically define the scope of archives open to the public.

[0003] At present, the file opening review work still adopts a fully manual review method, which requires human participation throughout the process, that is, it is necessary to manually review and verify the files one by one, and finally generate corresponding reports to determine which files can be opened, as well as the opening time and other information. This will undoubtedly lead to the consumption of a large amount of human resources. Summary of the invention

[0004] In order to improve the archive review work that consumes a lot of human resources, the embodiments of the present application provide an archive open review method, system and storage medium.

[0005] In a first aspect, this embodiment provides a method for file opening and review, the method comprising: Obtaining an audit work plan, and training a preset audit model based on the audit work plan to obtain a localized audit model; Receiving audit data, and determining target personnel to participate in the audit based on the audit data for security screening; Performing quality processing on the audit data to obtain digital audit data; Obtaining a preliminary review opinion based on the digital review data and the localized review model; After obtaining the preliminary review opinion, a review instruction is generated, and the review result corresponding to the review instruction is received; Based on the review result and the digital audit data, a data set is compiled, the localized audit model is trained using the data set to obtain an updated audit model, and it is determined whether the updated audit model meets the audit accuracy rate. If so, the updated audit model is determined as a localized audit model, and the review result is output; If not satisfied, a preset audit model is trained based on the audit work plan to obtain a localized audit model.

[0006] In some embodiments, receiving audit data includes: Obtaining the working state and network state of the processing end, determining whether the working state is an idle state and the network state is an online state, and if so, connecting to a data interface with the collection system, and receiving audit data through the data interface; If not, shut down the data interface with the collection system, and determine whether the working state is a busy state, and if not a busy state, generate an audit data receiving instruction to receive the audit data; If it is a busy state, continue to obtain the working state of the processing end, and when the working state is an idle state, generate an audit data receiving instruction to receive the audit data.

[0007] In some embodiments, the audit data includes at least one of daily audit data and confidential audit data, and the step of determining target personnel participating in the audit based on the audit data to conduct security screening includes: Determine whether the audit data only includes daily audit data, and if so, determine the regular staff corresponding to the daily audit data as the target personnel participating in the audit; If not, obtain the confidentiality level corresponding to the confidentiality audit data, and determine the staff corresponding to the confidentiality level and the regular staff corresponding to the daily audit data as target personnel participating in the audit.

[0008] In some embodiments, performing quality processing on the audit data to obtain digital audit data includes: Performing quality inspection on the audit data to generate a quality inspection report indicating that there are quality problems with the audit data; Cleaning the audit data based on the quality inspection report to correct the audit data to obtain cleaned audit data; The cleaning audit data is digitally processed to obtain digital audit data.

[0009] In some embodiments, determining whether the updated audit model satisfies the audit accuracy rate includes: Acquire a check sub-data set in the data set that is not involved in obtaining the update audit model, and split the check sub-data into check input data and check output data; Substituting the verification input data into the update audit model to obtain actual output data, and obtaining the similarity between the actual output data and the verification output data; Determine whether the similarity exceeds a preset similarity, and if so, the updated audit model meets the audit accuracy rate; If not, the updated audit model does not meet the audit accuracy rate.

[0010] In some of the embodiments, obtaining the audit work plan includes: Receive the file review task instruction, obtain the review requirements based on the file review task instruction, send the review requirements to the review team, and wait for receiving the review work plan when the review team completes the review work plan.

[0011] In some embodiments, the review result includes an automatic opening result and a delayed opening result, and the method further includes: Automatically making the automatic opening result public; Record the duration of generating the review result, determine whether the duration reaches the extension time corresponding to the extension result, and if so, make the extension result public; If not reached, continue to record the duration of generating the review result.

[0012] In a second aspect, this embodiment provides a file opening and review system, the system comprising: a model processing module, a data processing module and a review processing module; wherein: The model processing module is used to obtain an audit work plan, and train a preset audit model based on the audit work plan to obtain a localized audit model; The data processing module is used to receive audit data, determine the target personnel for parameter audit based on the audit data to perform security screening; and perform quality processing on the audit data to obtain digital audit data; The audit processing module is used to obtain a preliminary review opinion based on the digital audit data and the localized audit model; generate a review instruction after obtaining the initial opinion, receive a review result corresponding to the review instruction; and compile a data set based on the review result and the digital audit data; The model processing module is also used to use the data set to train the localized audit model to obtain an updated audit model, and determine whether the updated audit model meets the audit accuracy rate. If so, the updated audit model is determined as a localized audit model and the review result is output; if not, a preset audit model is trained based on the audit work plan to obtain a localized audit model.

[0013] In some of the embodiments, the system further comprises a preparation module; wherein, The preparation module is used to receive the archive review task instruction, obtain the review requirements based on the archive review task instruction, send the review requirements to the review team, and send the review work plan to the processing end when the review team completes the review work plan.

[0014] In a third aspect, this embodiment provides a computer-readable storage medium on which is stored a computer program that can be run on a processor, and when the computer program is executed by the processor, the archive opening review method described in the first aspect is implemented.

[0015] By adopting the above method, this application obtains the audit work plan, trains the preset audit model based on the audit work plan to obtain the localized audit model; receives the audit data, determines the target personnel involved in the audit based on the audit data for security screening; performs quality processing on the audit data to obtain digital audit data; obtains the preliminary review opinion based on the digital audit data and the localized audit model; generates a review instruction after obtaining the preliminary review opinion, and receives the review result corresponding to the review instruction; compiles a data set based on the review result and the digital audit data, uses the data set to train the localized audit model to obtain the updated audit model, and determines whether the updated audit model meets the audit accuracy rate. If it does, the updated audit model is determined as the localized audit model, and the review result is output; if it does not meet, the preset audit model is trained based on the audit work plan to obtain the localized audit model. The archive opening audit work is carried out based on the human-machine collaborative mode of "model + manual" to improve the archive audit work that consumes a lot of human resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flowchart of completing the file opening review work provided in the embodiment of the present application.

[0017] Figure 2 This is a block diagram of an archive opening review method provided in an embodiment of the present application.

[0018] Figure 3 It is a block diagram of receiving audit data provided by an embodiment of the present application.

[0019] Figure 4 This is a block diagram of performing quality processing on audit data to obtain digital audit data provided by an embodiment of the present application.

[0020] Figure 5 This is a connection diagram of an archive open review system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0021] In order to more clearly understand the purpose, technical solutions and advantages of the present application, the present application is described and illustrated below in conjunction with the accompanying drawings and embodiments. However, it should be understood by those of ordinary skill in the art that the present application can be implemented without these details. For those of ordinary skill in the art, it is obvious that various changes can be made to the embodiments disclosed in the present application, and without departing from the principles and scope of the present application, the general principles defined in the present application can be applied to other embodiments and application scenarios. Therefore, the present application is not limited to the embodiments shown, but conforms to the broadest scope consistent with the scope claimed for protection in the present application.

[0022] The embodiments of the present application are further described in detail below in conjunction with the drawings in the specification.

[0023] In order to improve the archive review work that consumes a lot of human resources, this application discloses a human-computer collaborative mode based on "model + manual" to carry out archive opening review work. Figure 1 This is a flow chart of completing the file opening and review work provided by the embodiment of the present application. Figure 1 As shown, in the process of archive opening review, there are nine steps: formulating the review work plan, customizing the localized review model, data handover, security screening, data quality processing, model-assisted review, professional review, model retraining and output of review results.

[0024] Figure 2 This is a block diagram of a method for file opening and review provided by an embodiment of the present application. Figure 2 As shown, a method for file opening review includes the following steps: Step S100, obtaining an audit work plan, and training a preset audit model based on the audit work plan to obtain a localized audit model.

[0025] This embodiment describes a method for file opening auditing from the perspective of the computer processing end that the model relies on. The above audit work plan is manually formulated and sent to the processing end so that the processing end can obtain the audit work plan. Among them, before obtaining the audit work plan, the processing end also includes: receiving the file audit task instruction, obtaining the audit requirements based on the file audit task instruction, sending the audit requirements to the audit team, and waiting to receive the audit work plan when the audit team completes the audit work plan.

[0026] When there is a need for archive opening and review, the staff responsible for archive management will send an archive review task instruction to the processing end, which represents a reminder of the expired archive review task, so that the processing end knows that it needs to prepare to carry out the archive opening and review work when it receives the archive review task instruction. In order to be able to carry out the archive opening and review work in a more consistent manner with the review requirements, the processing end will generate an audit requirement acquisition instruction when it receives the archive review task instruction, and send it to the relevant personnel, so that the relevant personnel will also send the audit requirements to the processing end, so that the processing end can obtain the audit requirements and send the audit requirements to the review team, so that the review team can combine the audit requirements for carrying out the archive opening and review work to clarify the audit work content and cycle, and study and discuss the progress of the audit data such as the age, as well as the on-site audit outbreak and review requirements, formulate an audit work plan, and clarify the audit work goals and tasks. When the audit team completes the audit work plan, it will send the audit work plan to the processing end so that the processing end receives the audit work plan, thereby completing the first link of formulating the audit work plan in the nine links. Among them, after the processing end sends the audit requirements to the audit team, it is in a state of waiting to receive the audit work plan, so that it can receive the audit work plan in time and provide an information basis for the smooth implementation of subsequent work.

[0027] After the processing end receives the audit work plan, it trains the preset audit model based on the audit work plan to obtain a localized audit model. Among them, the preset audit model is pre-trained by the relevant departments based on the keyword rules such as the local noun library, the responsible person thesaurus, the major event thesaurus, the policy thesaurus, the celebrity thesaurus, the sensitive event thesaurus, and the people's livelihood thesaurus. The preset audit model is a general audit model and is known. It is stored in the processing end and can be obtained by viewing the information stored in the processing end. The processing end substitutes the received audit work plan into the preset audit model, and then uses the local data in the local database to train the preset audit model substituted into the audit work plan. After training, a localized audit model suitable for localization is obtained. This can not only improve the supply capacity of the localized audit model, but also promote the efficient coupling and precise matching of the output supply side and the input demand side of the localized audit model, and optimize the model audit capacity. And complete the second link of formulating the localized audit model in the nine links to obtain the audit model.

[0028] Step S200, receiving audit data, and determining target personnel to participate in the audit based on the audit data for security screening.

[0029] After obtaining the audit model for the archive opening audit work, it is also necessary to determine the data that needs to be audited, that is, the processing end receives the audit data. Figure 3 is a block diagram of receiving audit data provided by an embodiment of the present application. Figure 3 As shown, receiving audit data includes the following steps: Step S201, obtain the working status and network status of the processing end, determine whether the working status is idle and the network status is normal, and if so, connect the data interface with the collection system to receive the audit data through the data interface.

[0030] Step S202: If not, shut down the data interface with the collection system and determine whether the working status is busy. If not, generate an audit data receiving instruction to receive the audit data.

[0031] Step S203, if it is a busy state, continue to obtain the working state of the processing end, and when the working state is an idle state, generate an audit data receiving instruction to receive the audit data.

[0032] The processing end and the collection system are provided with a data interface for transmitting data. The working status of the processing end includes an idle status and a busy status. The network status of the processing end includes an online status and an offline status.

[0033] Among them, the memory operation ratio of the processing end can be obtained by checking the current operation memory of the processing end at this time, dividing the current operation memory by the preset operation memory, and comparing the memory operation ratio with the preset operation ratio. If the memory operation ratio is greater than the preset operation ratio, the working state of the processing end is a busy state; if the memory operation ratio is not greater than the preset operation ratio, the working state of the processing end is an idle state. The preset operation ratio represents a value used to distinguish whether the processing end is in a busy state.

[0034] And you can check the network speed value of the processing end at this time and compare the network speed value with the value of zero. If the network speed value is zero, the network status of the processing end is disconnected; if the network speed value is not zero, the network status of the processing end is connected. Therefore, by checking the current running memory and network speed value and performing calculations and comparisons, the processing end can obtain the working status and network status of the processing end, and determine whether the working status is idle and the network status is connected.

[0035] When the processing end is determined to be idle and connected to the Internet, it indicates that the processor has the ability and conditions to receive the audit data online. At this time, the processing end will connect to the data interface with the collection system and receive the audit data sent by the collection system through the data interface. There is no need to send the audit data to the processing end offline, reducing manual workload and costs. Among them, the audit data is entry + full text. The full text import format supports PDF, JPG, TIFF, etc. The entry field value supports selective acquisition, such as file number, responsible person, year, file number, title, etc.

[0036] When it is determined that the processing end is not in an idle state and is in an online state, it indicates that the processing end does not have the ability and conditions to receive the audit data online at this time, or the processing end does not have the ability to receive the audit data online at this time, or the processing end does not have the conditions to receive the audit data online at this time, that is, the processing end cannot receive the audit data online at this time. At this time, in order to reduce the running memory caused by maintaining the connection with the collection system, the processing end will shut down the data interface with the collection system. In order to further determine the reason for the inability to receive the audit data online, it is also determined whether the working state is busy. If it is not busy, then the processing end is disconnected from the network at this time, and the processing end cannot receive the audit data online. The processing end will generate an audit data receiving instruction and send the audit data receiving instruction to the relevant personnel, so that the relevant personnel can manually send the audit data to the processing end so that the processing end can receive the audit data.

[0037] If the processing end is busy, then the processing end has no ability to receive audit data online regardless of whether it is connected to the Internet. In order to receive audit data when the processing end is not busy, the processing end generates an audit data receiving instruction when the working state is idle, and sends the audit data receiving instruction to the relevant personnel, so that the relevant personnel can manually send the audit data to the processing end, so that the processing end can receive the audit data. In this way, by combining the working state and network state of the processing end, the processing end will choose to receive audit data online only when it has the ability and conditions, and choose to receive audit data offline in other cases, so as to make full use of the operation and processing functions of the processing end to help reduce manual workload and costs. And complete the third link of data handover in the nine links to obtain audit data.

[0038] The audit data includes at least one of daily audit data and confidential audit data. The security screening based on the audit data to determine the target personnel for audit includes the following steps: Step S204, determining whether the audit data only includes routine audit data, and if so, determining the regular staff corresponding to the routine audit data as the target personnel to participate in the audit.

[0039] Step S205, if not, obtain the confidentiality level corresponding to the confidentiality audit data, and determine the staff corresponding to the confidentiality level and the regular staff corresponding to the daily audit data as the target personnel participating in the audit.

[0040] The processing end stores confidentiality keywords belonging to confidential audit data and the confidentiality level corresponding to each confidentiality keyword. After receiving the audit data, the processing end can compare the audit data with the confidentiality keywords. If the audit data contains the confidentiality keywords, then the audit data does not only contain daily audit data; if the audit data does not contain the confidentiality keywords, then the audit data only contains daily audit data.

[0041] It is determined that the audit data only includes daily audit data, indicating that undecrypted archival data will not be exposed during the archive opening audit work. At this time, regular staff only need to participate in the audit work, and there is no need for higher-level specific confidentiality staff to participate in the audit work, which can reduce the downgrading of staff. At this time, the regular staff corresponding to the daily audit data are the target personnel participating in the archive opening audit work for security screening.

[0042] The audit data does not only include daily audit data, indicating that during the archive opening audit work, undecrypted archive data will be exposed. In order to standardize the work, it is necessary to set the scope of personnel knowledge according to the principle of work needs and the principle of minimization, and the personnel with the right to know will conduct confidential screening, that is, by checking the confidentiality level corresponding to the confidentiality keyword to determine the confidentiality level corresponding to the confidentiality audit data, so that the staff corresponding to the confidentiality level will be responsible for the confidentiality screening of the audit data corresponding to the confidentiality level. For daily audit data that does not belong to confidential audit data, regular staff will conduct security screening. After all, it is impossible for all audit data for archive opening audit to be confidential audit data. In this way, when conducting security screening of audit data, measures such as data security and operation security can be implemented to ensure that the archive opening audit work is reasonable and standardized. And complete the fourth link of security screening in the nine links. Moreover, through the operation of security screening, the personnel involved in the subsequent review work can be determined.

[0043] Step S300, performing quality processing on the audit data to obtain digital audit data.

[0044] Because some archives were digitized at an early age, data quality issues in archive organization and digital processing directly affect the audit judgment. Therefore, data quality processing is essential before the audit. The audit data needs to be quality processed to obtain digital audit data. Figure 4 This is a block diagram of the quality processing of audit data to obtain digital audit data provided by the embodiment of the present application. Figure 4 As shown, performing quality processing on the audit data to obtain digital audit data includes the following steps: Step S301 , performing quality inspection on the audit data to generate a quality inspection report indicating that there are quality problems in the audit data.

[0045] Step S302, cleansing the audit data based on the quality inspection report to correct the audit data to obtain cleansed audit data.

[0046] Step S303, digitally process the cleaned audit data to obtain digital audit data.

[0047] Based on relevant national standards and specifications, the processing end can automatically perform quality inspection on the audit data with the help of archive data quality inspection tools to generate a quality inspection report that indicates quality problems in the audit data. The processing end then corrects and cleans the problems pointed out in the quality inspection report, such as the mismatch between the number of entries and the full text, the absence of full text, etc., to obtain cleaned audit data. Finally, combined with the characteristics of the digitized results of the cleaned audit data, the special OCR recognition technology needle is used to digitize the cleaned audit data to obtain digital audit data. Among them, if the audit data has no quality problems, the content of the quality inspection report is blank, and the cleaned audit data is the audit data. In this way, not only accurate digital audit data can be obtained, but also the format is changed to a data type that can be recognized by the localized audit model, which is convenient for the subsequent use of the localized audit model for archive opening audit work. At this time, the fifth link of data quality processing in the nine links is completed, that is, all preparations before the formal audit work are completed.

[0048] Step S400, obtaining a preliminary review opinion based on the digital review data and the localized review model.

[0049] The processing end builds a keyword rule base for opening and postponing opening, and assigns audit tasks for digital audit data, obtaining several tasks, each of which contains part of the digital audit data, and all tasks just constitute the audit tasks corresponding to all digital audit data. Then, the localized audit model is used to audit each task in turn. After all tasks are audited, the preliminary review opinion on the opening of this archive is obtained, that is, the audit opinion on opening or postponing opening is formed. In this way, by using the localized audit model instead of manual review, it can reduce manpower work and improve the archive review work that consumes a lot of human resources. At this time, the sixth link of the model-assisted audit in the nine links is completed.

[0050] Step S500, after obtaining the preliminary review opinion, a review instruction is generated, and a review result corresponding to the review instruction is received.

[0051] In order to ensure the accuracy of the audit, after obtaining the preliminary review opinion, the processing end also generates a review instruction, and sends the review instruction to the target personnel determined in the above step S200, so that these target personnel perform a manual review process, including manual review, unit review, archives final review and other steps. After the target personnel complete the review work, they will send the review results to the processing end, so that the processing end receives the review results. The preliminary review opinions are reviewed by professionals and data analysis is performed at the same time, thereby ensuring the accuracy of the audit. Moreover, only manual review of the preliminary review opinions is used, that is, only a review is performed without the need for preliminary review work, which can greatly reduce the workload and also play a role in improving the work of archive review that consumes a lot of human resources. At this time, the seventh link of professional review in the nine major links is completed.

[0052] Step S600, compile a data set based on the review results and the digitized audit data, use the data set to train the localized audit model to obtain an updated audit model, determine whether the updated audit model meets the audit accuracy rate, and if so, determine the updated audit model as a localized audit model and output the review results.

[0053] Step S700: If not satisfied, a preset audit model is trained based on the audit work plan to obtain a localized audit model.

[0054] The digital audit data and the corresponding review results are compiled into a data set according to the type of review results, and most of the data extracted from the data set is substituted into the localized audit model to train the localized audit model to obtain an updated audit model.

[0055] Among them, judging whether the updated audit model meets the audit accuracy rate includes the following steps: Step S601, obtaining the verification sub-data set in the data set that is not involved in obtaining the updated audit model, and splitting the verification sub-data into verification input data and verification output data.

[0056] Step S602: Substitute the verification input data into the update audit model to obtain the actual output data, and obtain the similarity between the actual output data and the verification output data.

[0057] Step S603, determine whether the similarity exceeds the preset similarity, and if so, update the audit model to meet the audit accuracy.

[0058] Step S604: If not, updating the audit model does not meet the audit accuracy rate.

[0059] The other data in the dataset that is not used to train the localized audit model is used as a validation set to test the accuracy of the update audit model. Among them, 70% of the data in the dataset can be selected for training, and the remaining 30% of the data can be used for verification. When verifying the accuracy of the update audit model, the similarity between the output of the update audit model and the result in the validation set is calculated, and the actual similarity obtained is compared with the preset similarity to obtain a conclusion on whether the update audit model meets the audit accuracy.

[0060] When the updated audit model meets the audit accuracy, it indicates that the model training is successful. At this time, the updated audit model is determined as a localized audit model, so that a more accurate audit model can be obtained in the subsequent audit. At this time, the eighth step of model retraining in the nine steps is completed.

[0061] Then the review results are output. The review results specifically include automatic opening results and delayed opening results. Specifically, the report is formed, and statistical reports such as "Statistical Table of Proposed Review Opinions for Each Year of Each Fond", "Summary Table of Proposed Open Catalogs", and "Archive Opening Review Form" are output according to the elements of the entire collection and year; the report is compiled, and professionals compile a comprehensive and detailed "Archive Opening Review Service Work Report" based on the analysis of the audit data; it is announced to the public, and in combination with the archives' requirements for public disclosure, an expiration reminder is set for the audited data, and it is automatically announced to the public after 25 years. In this way, the archive opening review work can be completed, which can improve the archive review work that consumes a lot of human resources. At this time, the ninth link of the audit results output in the nine links is completed.

[0062] When the updated audit model does not meet the audit accuracy, it indicates that the model training is unsuccessful, and then continue to execute step S100 to train the preset audit model based on the audit work plan to obtain a localized audit model. In this way, the model training result of step S600 feeds back to the model in step S100, thereby forming a closed-loop management of the archive opening audit work.

[0063] Preferably, this embodiment also includes automatically publishing the automatic opening result; recording the duration of generating the review result, judging whether the duration reaches the extension time corresponding to the extended opening result, if it reaches, publishing the extended opening result; if it does not reach, continuing to record the duration of generating the review result. This enables the processing end to automatically open the file in a targeted manner according to the review result, without the need to manually record the time to pay attention to when to open, reducing the workload of manpower, and also improving the file review work that consumes a lot of human resources.

[0064] Figure 5 Schematic diagram of the connection of an archive opening and review system provided in the embodiment of the present application. Figure 5As shown, an archive open review system includes a model processing module, a data processing module, a review processing module and a preparation module.

[0065] Among them, the model processing module is used to obtain the audit work plan, and train the preset audit model based on the audit work plan to obtain the localized audit model. The data processing module is used to receive the audit data, determine the target personnel for parameter audit based on the audit data to perform security screening; and perform quality processing on the audit data to obtain digital audit data. The audit processing module is used to obtain the initial review opinion based on the digital audit data and the localized audit model; generate the review instruction after obtaining the initial opinion, receive the review result corresponding to the review instruction; and compile the data set based on the review result and the digitalized audit data. The model processing module is also used to use the data set to train the localized audit model to obtain the updated audit model, determine whether the updated audit model meets the audit accuracy rate, if so, determine the updated audit model as the localized audit model, and output the review result; if not, train the preset audit model based on the audit work plan to obtain the localized audit model. The preparation module is used to receive the archive audit task instruction, obtain the audit requirements based on the archive audit task instruction, send the audit requirements to the audit team, and send the audit work plan to the processing end when the audit team completes the audit work plan.

[0066] The other functions performed by the above-mentioned model processing module, data processing module, review processing module and preparation module as well as the technical details of each function are the same or similar to the corresponding features in the archive opening review method described above, so they will not be repeated here.

[0067] An embodiment of the present application also provides a computer storage medium on which a computer program is stored. When the computer program is run on a computer, the computer can execute the steps of the archive opening and review method described above.

[0068] It should be understood that, although the steps in the flowchart of the accompanying drawings are shown in sequence according to the instructions of the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless explicitly stated in this document, the execution of these steps is not strictly limited in order and can be performed in other orders.

[0069] The above description is only a partial implementation method of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for file opening and review, characterized in that: The method comprises: Obtaining an audit work plan, and training a preset audit model based on the audit work plan to obtain a localized audit model; Receiving audit data, and determining target personnel to participate in the audit based on the audit data for security screening; Performing quality processing on the audit data to obtain digital audit data; Obtaining a preliminary review opinion based on the digital review data and the localized review model; After obtaining the preliminary review opinion, a review instruction is generated, and the review result corresponding to the review instruction is received; Compile a data set based on the review result and the digital audit data, use the data set to train the localized audit model to obtain an updated audit model, determine whether the updated audit model meets the audit accuracy rate, and if so, determine the updated audit model as a localized audit model, and output the review result; If not satisfied, a preset audit model is trained based on the audit work plan to obtain a localized audit model.

2. The method according to claim 1, characterized in that The received audit data includes: Obtaining the working state and network state of the processing end, determining whether the working state is an idle state and the network state is an online state, and if so, connecting to a data interface with the collection system, and receiving audit data through the data interface; If not, shut down the data interface with the collection system, and determine whether the working state is a busy state, and if not a busy state, generate an audit data receiving instruction to receive the audit data; If it is a busy state, continue to obtain the working state of the processing end, and when the working state is an idle state, generate an audit data receiving instruction to receive the audit data.

3. The method according to claim 2, characterized in that The audit data includes at least one of daily audit data and confidential audit data, and the step of determining the target personnel participating in the audit based on the audit data to conduct security screening includes: Determine whether the audit data only includes daily audit data, and if so, determine the regular staff corresponding to the daily audit data as the target personnel participating in the audit; If not, obtain the confidentiality level corresponding to the confidentiality audit data, and determine the staff corresponding to the confidentiality level and the regular staff corresponding to the daily audit data as target personnel participating in the audit.

4. The method according to claim 1, characterized in that: The performing quality processing on the audit data to obtain digital audit data comprises: Performing quality inspection on the audit data to generate a quality inspection report indicating that there are quality problems with the audit data; Cleaning the audit data based on the quality inspection report to correct the audit data to obtain cleaned audit data; The cleaning audit data is digitally processed to obtain digital audit data.

5. The method according to claim 1, characterized in that The determining whether the updated audit model meets the audit accuracy rate includes: Acquire a check sub-data set in the data set that is not involved in obtaining the update audit model, and split the check sub-data into check input data and check output data; Substituting the verification input data into the update audit model to obtain actual output data, and obtaining the similarity between the actual output data and the verification output data; Determine whether the similarity exceeds a preset similarity, and if so, the updated audit model meets the audit accuracy rate; If not, the updated audit model does not meet the audit accuracy rate.

6. The method according to claim 1, characterized in that Before obtaining an audit work plan include: Receive the file review task instruction, obtain the review requirements based on the file review task instruction, send the review requirements to the review team, and wait for receiving the review work plan when the review team completes the review work plan.

7. The method according to claim 1, characterized in that The review result includes an automatic opening result and a delayed opening result, and the method further includes: Automatically making the automatic opening result public; Record the duration of generating the review result, determine whether the duration reaches the extension time corresponding to the extension result, and if so, make the extension result public; If not reached, continue to record the duration of generating the review result.

8. A file opening and review system, characterized in that: The system includes: a model processing module, a data processing module and an audit processing module; wherein, The model processing module is used to obtain an audit work plan, and train a preset audit model based on the audit work plan to obtain a localized audit model; The data processing module is used to receive audit data, determine the target personnel for parameter audit based on the audit data to perform security screening; and perform quality processing on the audit data to obtain digital audit data; The audit processing module is used to obtain a preliminary review opinion based on the digital audit data and the localized audit model; generate a review instruction after obtaining the initial opinion, receive a review result corresponding to the review instruction; and compile a data set based on the review result and the digital audit data; The model processing module is also used to use the data set to train the localized audit model to obtain an updated audit model, and determine whether the updated audit model meets the audit accuracy rate. If so, the updated audit model is determined as a localized audit model and the review result is output; if not, a preset audit model is trained based on the audit work plan to obtain a localized audit model.

9. The system according to claim 8, characterized in that The system also includes a preparation module; wherein, The preparation module is used to receive the archive review task instruction, obtain the review requirements based on the archive review task instruction, send the review requirements to the review team, and send the review work plan to the processing end when the review team completes the review work plan.

10. A computer-readable storage medium having a computer program that can be run on a processor stored thereon, characterized in that: When the computer program is executed by the processor, the file opening and review method as described in any one of claims 1 to 7 is implemented.

Images