Artificial intelligence-based compliance management data processing method and system

Abstract

The invention discloses a compliance management data processing system and method based on artificial intelligence, and the method comprises the steps: S1, obtaining an auditing influence factor of a to-be-audited material, the auditing influence factor comprising basic information, a risk degree value and / or a first person experience value of the to-be-audited material, the basic information of the to-be-examined material comprises a bid invitation item type and a material content quantity to which the to-be-examined material belongs; and S2, after the audit influence factors of the to-be-audited material form first input data, inputting the first input data into the deep learning training model for processing to obtain a first audit duration corresponding to the to-be-audited material. Visibly, according to the scheme of the invention, the accuracy of determining the examination duration of the to-be-examined material is improved, and the risk degree value which is a consideration factor is introduced into the examination influence factors, so that the risk prevention and control degree and compliance of bid inviting work can be further improved and ensured.

Description

Technical Field

[0001] The present invention relates to compliance data processing technology, and in particular to a compliance management data processing method and system based on artificial intelligence. Background Art

[0002] Regarding the bidding and procurement management of enterprises, in order to make business processes more standardized and effectively achieve the goals of compliance risk prevention and control, cost reduction and efficiency improvement, and transformation, many enterprises currently use electronic bidding and procurement systems to realize the digital information construction of bidding and procurement compliance management.

[0003] The main function of the current electronic bidding and procurement system is to conduct intelligent detection and identification of the compliance of bidders. Once an abnormality is detected, the prompt mechanism will be triggered immediately. Among them, the intelligent detection function of compliance mainly includes: 1. Detection and identification of purchase records. If at least two purchase records involve the same IP address, the same contact name and / or the same contact phone number, the system will send an early warning message to the project management member (such as popping up a window for displaying the early warning signal on the management system interface and / or sending an early warning message to the mobile terminal held by the project management member) so that he should be informed and confirmed as soon as possible and implement the corresponding subsequent corrective measures; 2. Detection and identification of bidding materials. If the uploaded bidding materials are If the IP addresses involved are identical, the contact names of the uploaded and submitted documents are identical, the contact phone numbers are identical, the MAC addresses are identical, and / or the hard drive machine codes are identical, and / or the textual expressions of the same content in the bid materials (such as historical project performance or financial information) are very similar, the system will issue a warning message to alert the bidding personnel of non-compliant bidding practices (such as collusion or bid rigging). Third, the system identifies the bidders' equity information. The system can compare the bidders' corporate credit information, shareholder relationships, and executive relationships to identify any connections between bidders, such as controlling shareholders, participating shareholders, or shared executives, thereby preventing bid rigging and bid rigging. Therefore, the current electronic bidding and procurement systems primarily focus on intelligently detecting and identifying bid materials uploaded by bidders, enabling efficient and compliant processing of such materials. However, the functions provided by electronic bidding and procurement systems for the early stages of bidding, particularly the compilation and review of bidding materials, largely fail to consider the efficiency and compliance of these processes. At present, in the stage of compiling and reviewing the bidding materials, the system's main function is to assign the compiled bidding materials to the corresponding reviewers for review, and to compile and store the reviewed and revised bidding materials. As for the allocation of bidding material review tasks, in order to ensure that the review tasks can be completed with quality within the required time and to be able to reasonably allocate and arrange the workload of the reviewers, the system generally uses the review time required for the materials to be reviewed as the basis for the allocation of review tasks. However, the review time required for the materials to be reviewed is currently mainly determined by the historical work experience of the assigned personnel. Therefore, the determination of the review time is easily affected by subjective factors, resulting in low accuracy, which can easily lead to low efficiency in the overall review work and increase the workload of the reviewers. Summary of the Invention

[0004] This application aims to solve at least one of the technical problems existing in the prior art. To this end, this application proposes an artificial intelligence-based compliance management data processing method and system that can improve the accuracy of determining the review time of bidding materials.

[0005] In a first aspect, an embodiment of the present application provides a compliance management data processing method based on artificial intelligence, the method comprising the following steps:

[0006] S1. Obtaining review influencing factors of the materials to be reviewed, wherein the review influencing factors include basic information of the materials to be reviewed, a risk level value, and / or the experience value of the first person. The basic information of the materials to be reviewed includes the type of bidding project to which the materials to be reviewed belong and the amount of material content.

[0007] S2. After the review influencing factors of the materials to be reviewed are formed into the first input data, the first input data is input into the deep learning training model for processing to obtain the first review time corresponding to the materials to be reviewed.

[0008] In some embodiments, the risk level value is obtained by the following steps:

[0009] Obtain audit risk levels and / or personnel identity verification security levels;

[0010] The corresponding risk level value is determined based on the audit risk level and / or the personnel identity verification security level, wherein the audit risk level is determined based on the severity level corresponding to the audit risk triggering event, and the personnel identity verification security level is determined based on the security level corresponding to the personnel identity verification event.

[0011] In some embodiments, obtaining the audit risk level specifically includes:

[0012] In the event that an audit risk trigger event is detected, determine whether there is an audit risk trigger event of the highest severity level;

[0013] When it is determined that there is an audit risk trigger event of the highest severity level, the first level will be used as the audit risk level;

[0014] When it is determined that there is no audit risk triggering event of the highest severity level, a second level is determined according to the severity level corresponding to each audit risk triggering event, and the second level is used as the audit risk level.

[0015] In some embodiments, obtaining the personnel identity verification security level specifically includes:

[0016] Get the security level of each personnel authentication event;

[0017] When there is a personnel identity verification event of the highest security level, the third level shall be used as the personnel identity verification security level;

[0018] When there is no personnel identity authentication event of the highest security level, the fourth level is determined according to the security level of each personnel identity authentication event, and the fourth level is used as the personnel identity authentication security level.

[0019] In some embodiments, the first person's experience value is obtained by the following steps:

[0020] Acquire historical work information of the personnel, wherein the historical work information includes years of work and a first processing score of each historical project material;

[0021] Determine a first experience value based on a first processing score of each historical project material;

[0022] Determine the first weight coefficient based on years of work experience;

[0023] The first experience value is adjusted using the first weight coefficient to obtain a second experience value, and the second experience value is used as the first personnel experience value;

[0024] Therein, the working years are in positive proportion to the first weight coefficient, and the first weight coefficient and the first experience value are both in positive proportion to the first person's experience value.

[0025] In some embodiments, the first processing score of the historical project material is obtained by the following steps:

[0026] Obtaining a second weight coefficient corresponding to the historical project material and a second processing score for the historical project material, wherein the second weight coefficient is determined based on a similarity between the historical project material and the material to be reviewed;

[0027] A first processing score is determined based on the second weight coefficient and the second processing score.

[0028] In some embodiments, the second processing score of the historical project material is obtained by the following steps:

[0029] After obtaining the number of revisions of the historical project material, determining a first score value according to the number of revisions, wherein the number of revisions and the first score value are in inverse proportion;

[0030] Obtaining a third weight coefficient and / or a fourth weight coefficient, wherein the third weight coefficient is determined based on the modification duration of the historical project material, and the fourth weight coefficient is determined based on the importance of the modified content;

[0031] After adjusting the first rating value according to the third weight coefficient and / or the fourth weight coefficient, a second rating value is obtained.

[0032] In some embodiments, the review influencing factors also include the importance value of the material to be reviewed.

[0033] In some embodiments, the method further comprises the following steps:

[0034] S3. Display the review task of the materials to be reviewed on the first interface, wherein the first interface is provided with a first button, the first button is used to trigger the pop-up and display of the second interface, and the second interface is used to display and modify the review influencing factors and the first review duration.

[0035] In a second aspect, an embodiment of the present application provides an artificial intelligence-based compliance management data processing system, the system comprising:

[0036] A first acquisition unit is configured to acquire audit influencing factors of the materials to be reviewed, wherein the audit influencing factors include basic information of the materials to be reviewed, a risk level value, and / or the experience value of the first person, and the basic information of the materials to be reviewed includes the type of bidding project to which the materials to be reviewed belong and the amount of material content;

[0037] The first processing unit is used to form the first input data from the review influencing factors of the materials to be reviewed, and then input the first input data into the deep learning training model for processing to obtain the first review time corresponding to the materials to be reviewed.

[0038] The present application can achieve at least one of the following technical effects: the present application solution forms the first input data with the audit influencing factors of the materials to be reviewed, and then inputs it into the deep learning training model for processing to obtain the corresponding first audit time, wherein the audit influencing factors include the basic information of the materials to be reviewed, the risk level value and / or the first person's experience value, and the basic information of the materials to be reviewed includes the type of bidding project to which the materials to be reviewed belong and the amount of material content. It can be seen that the present application solution combines multi-dimensional considerations and deep learning training models to determine the audit time, which has high accuracy and can improve the processing efficiency and quality of the audit work. In addition, the risk level value is also introduced into the audit influencing factors, which can further improve the risk prevention and control level and compliance of the bidding work. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments.

[0040] Figure 1 A flowchart of a compliance management data processing method based on artificial intelligence is provided for an embodiment of the present application;

[0041] Figure 2Provided is a flowchart of the steps for determining the risk level value in an artificial intelligence-based compliance management data processing method according to an embodiment of the present application;

[0042] Figure 3 A flowchart of the steps for determining the first person's experience value in an artificial intelligence-based compliance management data processing method is provided for an embodiment of the present application;

[0043] Figure 4 A framework diagram of an artificial intelligence-based compliance management data processing system is provided for an embodiment of the present application. DETAILED DESCRIPTION

[0044] To make the objectives, technical solutions, and advantages of this application more clear, the following will refer to the drawings in the embodiments of this application to clearly and completely describe the technical solutions of this application through implementation methods. Obviously, the described embodiments are part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0045] In order to facilitate the execution of bidding and procurement management, more and more companies will adopt electronic bidding and procurement systems to realize the digital information construction of bidding and procurement compliance management, so as to play the role of compliance risk prevention and control, cost reduction and efficiency improvement. For the currently commonly used electronic bidding and procurement systems, their main function is to centrally carry out intelligent detection and identification of bidding materials uploaded by bidders, so as to realize efficient and compliant processing of bidding materials. As for the early stages of bidding, especially the stage of compiling and reviewing bidding materials, the corresponding functions of the system are only some basic functions, such as the distribution, compilation, and storage of review tasks for the bidding materials to be reviewed (which may at least include the bidding materials that the bidder needs to compile and provide during the bidding process, such as bidding announcements, bidding documents, and bidding invitations).

[0046] In currently commonly used electronic bidding and procurement systems, the review task of compiled bidding materials is generally based on the review time required for the bidding materials, thereby determining the review personnel assigned to the task, the completion deadline for the bidding materials review task, etc. Currently, the determination of the review time is basically based solely on the historical work experience of managers. This is not only easily affected by human subjective factors, but also does not take into account factors that may affect the review time, resulting in low accuracy in the determination of the review time and insufficient adaptation to actual conditions. In view of this, the embodiments of the present application provide a compliance management data processing system and method based on artificial intelligence, which can improve the accuracy of the determination of the review time.

[0047] Reference Figure 1, an embodiment of the present application provides a compliance management data processing method based on artificial intelligence, which includes the following steps.

[0048] S1. Obtain audit influencing factors of the materials to be reviewed, wherein the audit influencing factors include basic information of the materials to be reviewed, a risk level value and / or a first person's experience value, and the basic information of the materials to be reviewed includes the type of bidding project to which the materials to be reviewed belong and the amount of material content.

[0049] Specifically, materials awaiting review refer to bidding materials compiled by compilers and waiting to be sent to reviewers for review.

[0050] Audit influencing factors refer to factors that affect the audit time of materials to be reviewed. In the process of researching the plan, it was found that the audit influencing factors mainly include the following factors: 1. Basic information of materials to be reviewed - the basic information of materials to be reviewed mainly includes the type of bidding project to which the materials to be reviewed belong and the amount of material content. Among them, for different types of bidding projects, the corresponding difficulty is different. For example, the difficulty of the general contracting of construction projects is greater than that of the construction engineering construction bidding project, and the difficulty of the construction engineering construction bidding project is greater than that of the construction engineering supervision bidding project, or the difficulty of the construction engineering bidding project is greater than that of the service bidding project. The difficulty of the bidding project is greater, and so on. The more difficult the bidding project is, the longer the review time will be. Therefore, for the bidding project type to which the materials to be reviewed belong, its essence is to characterize the difficulty of reviewing the materials to be reviewed. Different bidding project types correspond to different review difficulties. The greater the difficulty of the bidding project, the greater the difficulty of the corresponding material review. As for the content of the materials to be reviewed, the more the content of the materials, the longer the review time will be. Among them, the content of the materials can be represented by parameters such as the total number of words and / or the number of pages of the materials. It can be seen that the basic information of the materials to be reviewed will affect the review time; 2. Risk level value - For the risk level value, its main It is determined based on the security risk events that occur during the compilation stage of the materials. Among them, the security risk events may include network security issues (such as whether the network is attacked), personnel identity security verification issues, material leakage issues, etc. When these security risk events occur during the compilation process of the materials, the auditors must review and check these security risk situations during the audit stage to determine whether there is any leakage of bidding material information. If so (that is, the higher the severity of the security risk event), it is necessary to further determine the content of the leakage and make corresponding modifications. It can be seen that if the risk level is higher, the auditors will need to spend longer time to perform the audit work, thus It can be seen that the risk level of security risk events that occur during the compilation stage by the compiler will affect the audit time of subsequent auditors; 3. The experience value of the first person - For the experience value of the first person, it refers to the compilation experience value of the compiler who compiles the materials to be reviewed. Among them, when the compiler has a relatively low number of years of work experience in the compilation of bidding materials, and / or his processing evaluation score in the compilation of bidding materials is low, then it means that the compiler has low experience in this regard, which will result in more audit revisions and / or the auditor will spend longer time reading and understanding the content of the materials. It can be seen that the experience value of the compiler will also affect the audit time.

[0051] S2. After the review influencing factors of the materials to be reviewed are formed into the first input data, the first input data is input into the deep learning training model for processing to obtain the first review time corresponding to the materials to be reviewed.

[0052] Specifically, the deep learning training model is a pre-trained deep learning model. The training steps for the deep learning training model may include: using the historical review influencing factors of the bidding materials as training input data, using the historical review duration of the bidding materials as training output data, and then training the deep learning training model using the training input data and training output data until the training termination conditions are met. The deep learning training model may be implemented using a multimodal fusion model, RNN, LSTM, or other existing models, depending on the actual situation and needs, and is not particularly limited here.

[0053] As can be seen from the above, the solution of the embodiment of the present application utilizes a combination of multi-dimensional considerations and deep learning training models in the method of determining the review time of the materials to be reviewed. This can greatly improve the accuracy of determining the review time, thereby improving the accuracy and rationality of the allocation of review tasks, and avoiding the situation where the review tasks cannot be completed within the specified time limit and / or the reviewers' tasks are particularly heavy due to unreasonable allocation. In addition, the risk level value is also introduced into the multi-dimensional considerations. In this way, the review time can not only meet the basic review modification requirements, but also meet the requirements of risk review and confirmation, thereby further improving the risk prevention and control level and compliance of bidding material processing.

[0054] In some embodiments, the risk level value is mainly determined based on the audit risk level and / or the personnel identity verification security level. The audit risk level is determined based on the severity level corresponding to the audit risk trigger event, and the audit risk trigger event mainly refers to the first security risk event that occurs when the compiler is compiling the materials, such as network attacks, improper operations of the compiler (such as taking pictures of the compiled materials with a mobile phone, using third-party equipment to access data, etc.), and other personnel who do not have the authority to view the bidding materials during the compilation process. They peek at the bidding materials in the compilation process, etc. These security risk events all have the problem of information leakage. Therefore, in addition to reviewing the materials during the audit process, these security risk events also need to be reviewed and checked to ensure that the materials currently under review have not been leaked. If leakage is found, the content needs to be rectified in a timely manner. Similarly, in addition to the security risk events that occur during the writing process, it is also necessary to consider whether identity authentication security risk issues occur when the editor logs into the system. For example, whether the editor uses the MAC address to log in to the system at a terminal in the specified area, the number of incorrect password entries exceeds the limit, the number of biometric identification errors exceeds the limit, and the identity authentication login information is peeped by others, etc. These problems will also cause information leakage. Therefore, when identity authentication security risk issues occur when logging into the system during the writing stage, they should also be reviewed and checked during the subsequent review process. It can be seen that compared to no security risk issues during the writing stage, when security risk issues occur during the writing stage, the review time will be longer, and the higher the severity of the security risk issue, the longer the review time will be. Therefore, referring to Figure 2 , the risk level value can be obtained through the following steps.

[0055] A1. Obtain audit risk level and / or personnel identity verification security level.

[0056] A2. Determine the corresponding risk level value based on the audit risk level and / or the personnel identity verification security level, wherein the audit risk level is determined based on the severity level corresponding to the audit risk triggering event, and the personnel identity verification security level is determined based on the security level corresponding to the personnel identity verification event.

[0057] Specifically, the audit risk level can be directly the severity level corresponding to the audit risk triggering event (i.e., the security risk event that occurs during the compilation of materials), or the audit risk level can be obtained after optimizing and adjusting the severity level. As for the personnel identity authentication security level, in the current identity authentication method, if a security risk problem occurs during identity authentication, then the identity authentication mechanism of the corresponding security level will be triggered according to the severity of the security risk problem. The more serious the security risk problem, the higher the security level of the identity authentication mechanism it triggers. Therefore, in this embodiment, the triggered personnel identity authentication security level is directly used to represent the severity of the security risk problem that occurs during identity authentication. It can be seen that the audit risk level and the personnel identity authentication security level are both in positive proportion to the risk level value. The higher the severity level and the personnel identity authentication security level corresponding to the audit risk triggering event, the greater the risk level value.

[0058] It can be seen that by adopting the above-mentioned method of determining the risk level value, it takes into account more comprehensively the situations in which security risk issues may arise, and while improving the accuracy of the audit time, it also further improves the risk prevention level and compliance of the audit work.

[0059] In some embodiments, the risk level value may directly include two variables, the audit risk level and the personnel identity verification security level, which are placed as characteristic parameters in the characteristic matrix of the first input data; alternatively, the risk level value is a variable, and then the final level value obtained by fusing the level values ​​of the audit risk level and the personnel identity verification security level is assigned to the risk level value. Given that the present application utilizes a deep learning training model to process the input feature data to obtain the audit duration, the risk level value is directly made to include the two variables, the audit risk level and the personnel identity verification security level, to constitute the first input data. This not only improves data processing efficiency but also achieves relatively high accuracy.

[0060] In some embodiments, since at least two audit risk triggering events may occur during the material compilation process, and the severity levels of these at least two audit risk triggering events may be the same or different, the specific implementation steps for obtaining the audit risk level may include the following steps.

[0061] A11. If it is detected that there is no audit risk triggering event, the audit risk level is set to 0; if it is detected that there is an audit risk triggering event, it is determined whether there is an audit risk triggering event of the highest severity level.

[0062] A12. When it is determined that there is an audit risk trigger event with the highest severity level, the first level will be used as the audit risk level.

[0063] Specifically, among several audit risk triggering events, if at least one audit risk trigger of the highest severity level occurs, then the audit risk level is set to the first level, and the first level is used to represent the highest audit risk level. If the smaller the audit risk level value, the higher the severity level, then the first level value is the smallest, and its value is smaller than the second level value. Conversely, if the larger the audit risk level value, the higher the severity level of the audit risk triggering event, then the first level value is the largest, and its value is larger than the second level value. Furthermore, in this embodiment, the larger the audit risk level value, the higher the severity level of the audit risk triggering event. In this way, the highest severity level of the audit risk triggering event can be directly used as the first level. For example, if the highest severity level is level 5, then the audit risk level can also be set to 5. Of course, the first level can also be obtained by adjusting the highest severity level by adding a weighting coefficient (which is an empirical value), such as first level = weighting coefficient k * highest severity level. This can be selected and set according to actual needs and is not specifically limited here.

[0064] A13. When it is determined that there is no audit risk trigger event with the highest severity level, the second level is determined based on the severity level corresponding to each audit risk trigger event, and the second level is used as the audit risk level.

[0065] Specifically, the step of determining the second level according to the severity level corresponding to each audit risk triggering event specifically includes: A131, obtaining the severity levels {y1, y2, y3, ... y n}, where y1 represents the severity level corresponding to the first audit risk trigger event, and so on; A132, after calculating the average severity level corresponding to several audit risk trigger events, the average severity level Y is obtained. avg =(y1+y2+y 3+ …+y n ) / n, and then take the average severity level as the second level, where n represents the total number of severity levels corresponding to the audit risk trigger events. Of course, if the weighting coefficient k1 is introduced in the method of determining the audit risk level based on the severity level, then for the second level, it should be k1*Y avg , thereby ensuring that the first level is necessarily greater than the second level. It should also be noted that the setting of audit risk trigger events and their corresponding severity levels can be implemented using the existing risk prevention and control functions in the electronic bidding and procurement system, and no specific limitations are made here.

[0066] In some embodiments, among various audit risk triggering events, the recognition accuracy of the event of material peeping behavior is relatively low. The current recognition method is to use the camera set on the computer terminal used by the editor (the camera is set in front of the display screen, that is, on the same side as the display interface, and can be used to shoot the area currently in front of the display interface) to shoot the area in front of the display interface, thereby identifying whether there are other people who do not have the right to view the materials staying in the area in front of the display interface. If so, it is determined that there is material peeping behavior, otherwise, it is determined that there is no material peeping behavior. However, this will result in the situation where the person only passes by and stays in the area, but the line of sight does not stay on the display interface, which will also be determined as material peeping behavior. Therefore, in order to further improve the recognition accuracy of the audit risk triggering event of material peeping behavior and its corresponding severity level, the audit risk triggering event can be determined through the following steps.

[0067] A01. Obtain a first video image; wherein, the first video image is obtained by shooting with a camera arranged in front of the display screen, that is, the camera is mainly used to shoot the area in front of the display screen.

[0068] A02. When the display interface of the display screen displays the content of the materials to be reviewed, face recognition is performed on the first video image to obtain a first face image.

[0069] A03. After performing personal identity recognition on the face in the first facial image, obtain the total authority corresponding to the personal identity.

[0070] Specifically, after a face recognition algorithm is used to identify the face in the first face image, the identity of the person corresponding to the face is obtained. Then, the total authority corresponding to the identity of the person is obtained from a preset database. The face recognition algorithm can be implemented using an existing face identity verification algorithm, which is not elaborated on here.

[0071] A04. When it is determined based on the total authority that the person does not have the first authority, the eye area of ​​the first facial image is located and identified; when the eye area is identified, the audit risk triggering event is confirmed and processed.

[0072] Specifically, when peeking at the contents of the materials to be reviewed, the person's face must be facing the display interface. In this case, the eye region image can be located from the first facial image using eye features. Conversely, if the eye region image cannot be identified and located from the first facial image, it indicates that the person has not actually viewed the contents of the materials to be reviewed on the display interface. This indicates that eye region location can prevent false triggering of audit security risk events due to a person simply lingering in front of the display screen without viewing the materials, thereby greatly improving the accuracy of audit risk trigger event identification.

[0073] Furthermore, to improve the accuracy of determining the audit risk trigger event, pupil location can be performed on the eye area image, and then the pupil's gaze direction can be estimated to determine whether the person's gaze falls on the materials for review displayed on the display interface, or even determine whether their gaze falls on the display interface. This can avoid the false triggering of the audit risk trigger event due to the person's eyes pointing toward the display interface but not actually falling on the materials for review. Therefore, step A04 specifically includes the following steps.

[0074] A041. When it is determined based on the total authority that the person does not have the first authority, perform eye area positioning and identification on the first facial image.

[0075] A042. When the eye area is identified, the pupil area is obtained from the eye area.

[0076] A043. After calculating the pupil sight direction of the pupil area using a sight direction estimation algorithm, a pupil sight direction vector is obtained.

[0077] A044. Based on the coordinate conversion mapping relationship between the eye parameter coordinate system and the display screen coordinate system, the pupil sight direction vector in the eye parameter coordinate system is converted into the first coordinate vector in the display screen coordinate system. Then, based on the position of the first coordinate vector in the display screen coordinate system, it is determined whether the person's pupil sight falls on the page of the material to be reviewed.

[0078] A045. When it is determined that the person's pupil gaze falls on the page of materials to be reviewed, it is determined that there is an audit risk trigger event.

[0079] The severity level corresponding to the audit risk triggering event can be determined through the following steps.

[0080] A05. When it is determined that an audit security risk event exists, obtain the currently displayed page content, wherein the currently displayed page content refers to the content of the pending review materials currently displayed on the display interface.

[0081] A06. Identify the importance level of the page content and determine the severity level corresponding to the audit risk trigger event based on the identified importance level. The higher the identified importance level, the higher the severity level.

[0082] Specifically, since the importance of technical specifications and financial content is higher than that of procedural content, and the importance of key indicator parameters is higher than that of procedural / routine indicator parameters, the adverse effects caused by the leakage of technical specifications and financial content before the announcement will be greater than that of procedural content, or the leakage of key indicator parameters before the announcement will be greater than that of procedural / routine indicator parameters. Therefore, in the case of the same audit risk triggering event, if the importance of the peeped material is greater, the corresponding severity level of the event should be higher.

[0083] It can be seen that by adopting the above method to determine the audit risk triggering events and their corresponding severity levels, its accuracy is higher and the situation of false triggering can be greatly reduced, thereby avoiding the increase of ineffective work of staff and improving the overall work processing efficiency.

[0084] In addition, it should be noted that for the image positioning and recognition of the above-mentioned pupil area, it can be implemented by using existing algorithms such as ellipse fitting and / or CNN convolutional neural network; similarly, the gaze direction estimation algorithm can be implemented by using the gaze vector settlement algorithm, which are not elaborated in detail here.

[0085] In some embodiments, in view of the above-mentioned method of obtaining the audit risk level, similarly, the step of obtaining the personnel identity authentication security level may specifically include the following steps.

[0086] A14. Obtain the security level of each personnel identity verification event.

[0087] A15. When there is a personnel identity authentication event of the highest security level, the third level shall be used as the personnel identity authentication security level.

[0088] A16. When there is no personnel identity authentication event of the highest security level, a fourth level is determined based on the security level of each personnel identity authentication event, and the fourth level is used as the personnel identity authentication security level.

[0089] Specifically, if there are five security levels for personnel identity authentication events, where 5 represents the highest security level, then when there is a personnel identity authentication event with the highest security level, 5 is used as the personnel identity authentication security level; otherwise, the security levels of several personnel identity authentication events, such as 2, 3, 3, and 4, are averaged to obtain the security level average value S. avg =(s1+s2+s3+ …+s m ) / m, and then the average security level is used as the fourth level; where s1 represents the security level of the first personnel identity authentication event, and so on, and m represents the total number of security levels of personnel identity authentication events. Of course, if a weighting coefficient k2 is introduced in the method of determining the personnel identity authentication security level based on the security level, then for the fourth level, it should be k2*Savg, thereby ensuring that the third level is greater than the fourth level. It should also be noted that the setting of personnel identity authentication events and their corresponding security levels can be implemented through the existing identity authentication security mechanism in the electronic bidding and procurement system, and will not be elaborated on here.

[0090] It can be seen that by adopting the above-mentioned method of obtaining the audit risk level and personnel identity verification security level, the accuracy of determining the risk level value can be further improved, providing accurate data support for the determination of subsequent audit time.

[0091] In some embodiments, reference Figure 3 , the first person's experience value is obtained through the following steps.

[0092] B1. Obtain the personnel's historical work information, wherein the historical work information includes years of work and a first processing score for each historical project material.

[0093] Specifically, the years of work experience mainly refers to the years of work experience related to the processing of bidding materials, while the first processing score of historical project materials is mainly used to characterize the quality of the work of the compiler in each previous processing of the bidding materials content.

[0094] B2. Determine a first experience value based on the first processing score of each historical project material.

[0095] Specifically, in this embodiment, the specific calculation method of the first experience value is: first experience value = (p1+p2+p3+…+p j ) / j, where p represents the first processing score of the historical project material, p i represents the first processing score of the i-th historical project material, i.e., p1 represents the first processing score of the first historical project material, and so on; j represents the total number of processing scores for historical project materials. In other words, step B2 specifically involves calculating the average of the first processing scores of several historical project materials to obtain a first score average, which is then used as the first empirical value.

[0096] B3. Determine a first weight coefficient based on years of work experience, wherein the years of work experience and the first weight coefficient are in positive proportion.

[0097] B4. Use the first weight coefficient to adjust the first experience value to obtain a second experience value, and use the second experience value as the first person's experience value; the first weight coefficient and the first experience value are both in positive proportion to the first person's experience value.

[0098] Specifically, to improve the accuracy of the first experience value and more accurately reflect the compiler's experience level, it is preferred to adjust the first experience value using a weight coefficient corresponding to years of work experience. Specifically, for the second experience value, the formula is: second experience value = first weight coefficient k3 * first experience value. In this case, the second experience value is the first person's experience value.

[0099] In some embodiments, due to differences between the project type, details, and other information of the historical project materials being processed and the materials to be reviewed, in order to further improve the accuracy of the first experience value, the processing score corresponding to the historical project materials that are more similar to the materials to be reviewed is given a larger weight. In other words, the more similar the historical project materials are to the materials to be reviewed, the more their corresponding processing score reflects the compiler's experience in compiling the materials to be reviewed. In view of this, the first processing score of the historical project materials can be obtained through the following steps.

[0100] C1. Obtain a second weight coefficient corresponding to the historical project material and a second processing score for the historical project material, wherein the second weight coefficient is determined based on a similarity between the historical project material and the material to be reviewed.

[0101] C2. Determine a first processing score based on the second weight coefficient and the second processing score.

[0102] Specifically, the second processing score of the historical project material is the original processing score. Then, the higher the similarity between the historical project material and the pending material, the higher the influence of the corresponding original processing score should be. Therefore, the similarity and the second weight coefficient are in a positive proportional relationship. The greater the similarity, the greater the second weight coefficient. Then, the first processing score = the second weight coefficient k4 * the second processing score. Then, the processing score obtained after processing, that is, the first processing score, is used as the final required processing score. In addition, for the similarity between materials, it can be achieved using existing algorithms such as algorithms based on TF-IDF and cosine similarity, algorithms based on BERT semantic embedding, etc., which will not be elaborated here.

[0103] In view of the introduction of the above similarity, the step B2 may specifically include the following steps.

[0104] B21. Obtain the similarity between each historical project material and the material to be reviewed. At this point, several similarities are obtained.

[0105] B22. Filter out a first similarity from the plurality of similarities, wherein the first similarity is a similarity greater than or equal to a first threshold, the second processing score of the historical project material corresponding to the first similarity is a third processing score, and the second similarity is a similarity less than the first threshold.

[0106] B23. When at least one first degree of similarity exists, the second processing score for the historical project material corresponding to the second degree of similarity is deleted, retaining the third processing score. A first empirical value is then determined based on the retained third processing score. This means that after averaging several third processing scores, the second score average is obtained, which is then used as the first empirical value. In other words, the retained third processing score can now be understood as the first processing score. This indicates that in determining the first empirical value, the processing scores corresponding to historical project materials that are more similar to the materials to be reviewed are used as the basis, while the processing scores corresponding to historical project materials with less similarity to the materials to be reviewed are not considered.

[0107] B24. If the first similarity does not exist, a first processing score is determined based on the second weight coefficient and the second processing score, and then a first experience value is determined based on the first processing score for each historical project material. The second weight coefficient and the second processing score are as described above and are not further elaborated here.

[0108] It can be seen that by using the above method to determine the first experience value, it can better reflect the experience level of the compiler in processing the materials to be reviewed, which can further provide more accurate data support for the determination of the subsequent review time.

[0109] In some embodiments, in order to further improve the accuracy of the first person's experience value, the second processing score of the historical project material is obtained through the following steps.

[0110] E1. After obtaining the number of revisions to the historical project materials, determine a first score based on the number of revisions. The number of revisions and the first score are inversely proportional. Specifically, the number of revisions refers to the total number of times the historical project materials were revised from submission to review until the review was completed. Therefore, the greater the number of revisions, the lower the first score.

[0111] E2. Obtain a third weight coefficient and / or a fourth weight coefficient, wherein the third weight coefficient is determined based on the modification duration of the historical project materials, and the fourth weight coefficient is determined based on the importance of the modified content.

[0112] Specifically, the modification time of historical project materials refers to the modification time required for a manuscript return, and the method for determining the modification time can be: the moment when the editor receives the manuscript return is taken as the starting time, and the moment when the manuscript return is submitted to the reviewer is taken as the ending time, and then the duration between the starting time and the ending time is taken as the modification time. Furthermore, in order to improve the accuracy of the determination of this modification time, the moment when the manuscript return is first opened by the editor can be taken as the starting time. Therefore, the modification time of a historical project material can include T1, T2, T3, ..., T l , where l represents the total number of revisions of the historical project material, and T1 represents the time spent on the first revision, i.e., the first revision time, T2, T3, ..., T l The third weight coefficient k5 is determined by calculating the average modification duration of several historical project materials to obtain the average duration, and then determining the third weight coefficient k5 based on the average duration. The average duration and the third weight coefficient are inversely proportional, that is, the shorter the average modification duration, the larger the third weight coefficient.

[0113] Regarding the importance of the modified content, the reviewers will generally mark the places in the materials that need to be modified in the form of comments and / or revisions. Therefore, the comments and / or revisions in the returned manuscripts can be identified, and then the importance of the content can be determined according to the type of the identified content. For example, the importance of technical specifications and financial contents is higher than that of procedural contents. Therefore, in the review process of historical project materials, there are several importance levels of modified content, such as V1, V2, V3, ..., V4, V5, V6, V7, V8, V9, V10, V11, V12, V13, V14, V15, V16, V17, V18, V19, V20, V21, V30, V41, V52, V63, V74, V85, V96, V11, V12, V13, V14, V15, V16 p , where p represents the total number of importance level values ​​of the modified content, V1 represents the importance level value of the first modified content, V2, V3, ..., V p The fourth weight coefficient k6 is determined by calculating the average importance level of several modifications to the historical project materials to obtain the average importance level. The fourth weight coefficient k6 is then determined based on the average importance level. The average importance level and the fourth weight coefficient k6 are inversely proportional, i.e., the greater the average importance level of the modification, the smaller the fourth weight coefficient k6.

[0114] E3. After adjusting the first score value based on the third weight coefficient and / or the fourth weight coefficient, a second score value is obtained. The third weight coefficient, the fourth weight coefficient, and the first score value are all directly proportional to the second score value. In this embodiment, the second score value is specifically: k5 * k6 * first score value. The calculation formula for the second score value may optionally include an empirical fixed value for fine-tuning to achieve greater accuracy. This can be determined based on actual needs and is not specifically limited herein.

[0115] In some embodiments, the audit influencing factors also include the importance of the materials to be reviewed. The importance of the materials to be reviewed is generally manually set, mainly for situations where the bidding projects to be reviewed are of the same type and other factors are relatively similar. In this case, it is necessary for management personnel to determine.

[0116] In some embodiments, in view of the fact that before the audit task of the materials to be reviewed is assigned, there may be a situation where there is a strong human intervention in the setting of the audit influencing factors and / or the first audit duration, a function that can modify the audit task of the materials to be reviewed is set on the first interface for displaying the audit task of the materials to be reviewed. Therefore, for the method of this application, it may also include the following steps: S3, the audit task of the materials to be reviewed is displayed on the first interface, wherein the first interface is provided with a first button, the first button is used to trigger the second interface to pop up and display, and the second interface is used to display and modify the audit influencing factors and the first audit duration. This can help the staff to modify and determine the audit influencing factors and / or the first audit duration. Of course, for this modification function, it needs to be authenticated before it is allowed to be started, so as to avoid modifications by unauthorized staff, thereby improving the security and reliability of system operation.

[0117] Reference Figure 4 , an embodiment of the present application provides an artificial intelligence-based compliance management data processing system, the system comprising:

[0118] A first acquisition unit is configured to acquire audit influencing factors of the materials to be reviewed, wherein the audit influencing factors include basic information of the materials to be reviewed, a risk level value, and / or the experience value of the first person, and the basic information of the materials to be reviewed includes the type of bidding project to which the materials to be reviewed belong and the amount of material content;

[0119] The first processing unit is used to form the first input data from the review influencing factors of the materials to be reviewed, and then input the first input data into the deep learning training model for processing to obtain the first review time corresponding to the materials to be reviewed.

[0120] The units of the above system embodiment correspond one to one with the steps of the method embodiment, so the implementation principle and beneficial effects of the system embodiment are the same as those of the above method embodiment, and will not be repeated here.

[0121] In addition, an embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program is executed by a processor to implement the steps of the above method embodiment.

[0122] For the processors mentioned in the above storage medium embodiment and system embodiment, the number can be at least one, and at least any step in the above method embodiment can be executed. When the number is at least two, at least two processors can be connected to each other for communication, not limited to wired or wireless communication connection, and the at least one processor can be connected to various intelligent terminal devices for communication. In addition, the processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0123] Finally, it should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0124] Note that the above are only preferred embodiments of the present application and the technical principles employed. Those skilled in the art will understand that the present application is not limited to the specific embodiments described herein, and that various obvious changes, readjustments, and substitutions can be made by those skilled in the art without departing from the scope of protection of the present application. Therefore, although the present application has been described in more detail through the above embodiments, the present application is not limited to the above embodiments and may include many other equivalent embodiments without departing from the scope of the present application. The scope of the present application is determined by the scope of the appended claims.

Claims

1. A compliance management data processing method based on artificial intelligence, characterized in that: The method comprises the following steps: S1. Obtaining review influencing factors of the materials to be reviewed, wherein the review influencing factors include basic information of the materials to be reviewed, a risk level value, and / or the experience value of the first person. The basic information of the materials to be reviewed includes the type of bidding project to which the materials to be reviewed belong and the amount of material content. S2. After the review influencing factors of the materials to be reviewed are formed into the first input data, the first input data is input into the deep learning training model for processing to obtain the first review time corresponding to the materials to be reviewed.

2. The method according to claim 1, wherein The risk level value is obtained by the following steps: Obtain audit risk levels and / or personnel identity verification security levels; The corresponding risk level value is determined based on the audit risk level and / or the personnel identity verification security level, wherein the audit risk level is determined based on the severity level corresponding to the audit risk triggering event, and the personnel identity verification security level is determined based on the security level corresponding to the personnel identity verification event.

3. The method according to claim 2, wherein The acquisition and audit risk level specifically includes: In the event that an audit risk trigger event is detected, determine whether there is an audit risk trigger event of the highest severity level; When it is determined that there is an audit risk trigger event of the highest severity level, the first level will be used as the audit risk level; When it is determined that there is no audit risk triggering event of the highest severity level, a second level is determined according to the severity level corresponding to each audit risk triggering event, and the second level is used as the audit risk level.

4. The method according to claim 2, wherein The security level of obtaining the identity verification of the personnel specifically includes: Get the security level of each personnel authentication event; When there is a personnel identity verification event of the highest security level, the third level shall be used as the personnel identity verification security level; When there is no personnel identity authentication event of the highest security level, the fourth level is determined according to the security level of each personnel identity authentication event, and the fourth level is used as the personnel identity authentication security level.

5. The method according to claim 1, wherein The first personnel experience value is obtained by the following steps: Acquire historical work information of the personnel, wherein the historical work information includes years of work and a first processing score of each historical project material; Determine a first experience value based on a first processing score of each historical project material; Determine the first weight coefficient based on years of work experience; The first experience value is adjusted using the first weight coefficient to obtain a second experience value, and the second experience value is used as the first personnel experience value; Therein, the working years are in positive proportion to the first weight coefficient, and the first weight coefficient and the first experience value are both in positive proportion to the first person's experience value.

6. The method according to claim 5, wherein The first processing score of the historical project material is obtained by the following steps: Obtaining a second weight coefficient corresponding to the historical project material and a second processing score for the historical project material, wherein the second weight coefficient is determined based on a similarity between the historical project material and the material to be reviewed; A first processing score is determined based on the second weight coefficient and the second processing score.

7. The method according to claim 6, wherein The second processing score of the historical project material is obtained by the following steps: After obtaining the number of revisions of the historical project material, determining a first score value according to the number of revisions, wherein the number of revisions and the first score value are in inverse proportion; Obtaining a third weight coefficient and / or a fourth weight coefficient, wherein the third weight coefficient is determined based on the modification duration of the historical project material, and the fourth weight coefficient is determined based on the importance of the modified content; After adjusting the first rating value according to the third weight coefficient and / or the fourth weight coefficient, a second rating value is obtained.

8. The method according to claim 1, wherein The audit influencing factors also include the importance value of the materials to be reviewed.

9. The method according to any one of claims 1 to 8, wherein The method further comprises the following steps: S3. Display the review task of the materials to be reviewed on the first interface, wherein the first interface is provided with a first button, the first button is used to trigger the pop-up and display of the second interface, and the second interface is used to display and modify the review influencing factors and the first review duration.

10. An artificial intelligence-based compliance management data processing system, characterized in that: The system includes: A first acquisition unit is configured to acquire audit influencing factors of the materials to be reviewed, wherein the audit influencing factors include basic information of the materials to be reviewed, a risk level value, and / or the experience value of the first person, and the basic information of the materials to be reviewed includes the type of bidding project to which the materials to be reviewed belong and the amount of material content; The first processing unit is used to form the first input data from the review influencing factors of the materials to be reviewed, and then input the first input data into the deep learning training model for processing to obtain the first review time corresponding to the materials to be reviewed.

Images