Intelligent account closing method and system

Abstract

The invention provides an intelligent account closing method and system, and relates to the technical field of data processing, and the method comprises the steps: building a historical account closing database, carrying out the data feature extraction of the historical account closing database, synchronizing a data feature extraction result to a prediction model, generating predicted account closing time, and focusing on account closing abnormity; configuring a white list user during the account closing period, and establishing a white list backtracking window; performing account closing processing based on the financial module, releasing the white list users during the account closing period, and performing real-time linkage by using the financial module to complete the account closing processing; behavior monitoring is carried out on the white list user during the account closing period, and real-time early warning is generated based on a white list backtracking window and a behavior monitoring result; according to the technical scheme, intelligent account closing management is carried out according to real-time early warning, the technical problem that in the prior art, intelligent account closing cannot be carried out according to requirements, and consequently working efficiency is low is solved, intelligent account closing management is carried out by configuring white list users, and the technical effects of improving working efficiency and reducing business risks are achieved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to an intelligent account closing method and system. Background Art

[0002] In special periods such as the end of the year or the end of the month in hospitals, in order to conduct an inventory of accounts, it is usually necessary to shut down the business system to ensure the integrity and accuracy of the financial data. However, most hospitals currently still use the method of manually shutting down the business system to handle this. Due to the large number of business systems involving financial data, the manual shutdown process is not only labor-intensive, but also prone to omissions or operational errors. In addition, after the system is shut down, some users may need to temporarily use the system in specific scenarios, which requires the system to be temporarily opened manually, which is not only complicated and inconvenient to operate, but also prone to system configuration errors due to human factors, bringing potential risks. The traditional manual closing mode cannot automatically control the time of system opening and closing according to actual needs, resulting in low work efficiency and unable to meet the hospital's needs for rapid response and efficient processing of account inventory work during special periods. Summary of the invention

[0003] The present application provides an intelligent account closing method and system, which solves the technical problem in the prior art that intelligent account closing cannot be performed according to demand, resulting in low work efficiency.

[0004] In view of the above problems, the present application provides an intelligent account closing method and system.

[0005] In a first aspect, the present application provides an intelligent account closing method, the method comprising: establishing a historical account closing database, extracting data features from the historical account closing database, synchronizing the data feature extraction results to a prediction model, generating a predicted account closing time and concerned account closing anomalies; configuring whitelist users during the account closing period, and establishing a whitelist backtracking window; performing account closing processing based on a financial module according to the predicted account closing time, the concerned account closing anomalies, and a closing list, and releasing the whitelist users during the account closing period, the closing list comprising a budget module, a cost module, a purchase and sales module, and a reimbursement module, and completing the account closing processing after the budget module, the cost module, the purchase and sales module, and the reimbursement module are linked in real time by using the financial module; monitoring the behavior of the whitelist users during the account closing period, and generating a real-time warning based on the whitelist backtracking window and the behavior monitoring results; and performing intelligent account closing management according to the real-time warning.

[0006] In a second aspect, the present application provides an intelligent account closing system, the system comprising: a database establishment unit: establishing a historical account closing database, extracting data features from the historical account closing database, synchronizing the data feature extraction results to a prediction model, and generating a predicted account closing time and a concerned account closing anomaly; a whitelist configuration unit: configuring whitelist users during the account closing period, and establishing a whitelist backtracking window; an account closing processing unit: performing account closing processing based on a financial module according to the predicted account closing time, the concerned account closing anomaly, and a closing list, and releasing the whitelist users during the account closing period, the closing list comprising a budget module, a cost module, a purchase and sales inventory module, and a reimbursement module, and completing the account closing processing after the budget module, the cost module, the purchase and sales inventory module, and the reimbursement module are linked in real time using the financial module; a behavior monitoring unit: monitoring the behavior of the whitelist users during the account closing period, and generating a real-time warning based on the whitelist backtracking window and the behavior monitoring results; and an account closing management unit: performing intelligent account closing management according to the real-time warning.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] First, a historical account closing database is established. After extracting data features from the historical account closing database, the data feature extraction results are synchronized to the prediction model to generate the predicted account closing time and the account closing anomalies to be paid attention to. Next, whitelist users during the account closing period are configured, and a whitelist backtracking window is established. Then, account closing processing based on the financial module is performed according to the predicted account closing time, the account closing anomalies to be paid attention to, and the account closing list, and whitelist users are released during the account closing period. The account closing list includes the budget module, the cost module, the purchase and sales module, and the reimbursement module. The budget module, the cost module, the purchase and sales module, and the reimbursement module are linked in real time by the financial module to complete the account closing processing. Furthermore, the behavior of whitelist users during the account closing period is monitored, and real-time warnings are generated based on the whitelist backtracking window and the behavior monitoring results. Finally, intelligent account closing management is performed based on the real-time warning. The technical problem that the existing technology cannot perform intelligent account closing according to demand, resulting in low work efficiency, is solved. By configuring whitelist users for intelligent account closing management, the technical effect of improving work efficiency and reducing business risks is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 A flowchart of an intelligent account closing method is provided for this application;

[0010] Figure 2 A structural diagram of an intelligent account closing system is provided for this application.

[0011] Description of the accompanying drawings: database establishment unit 11, whitelist configuration unit 12, account closing processing unit 13, behavior monitoring unit 14, account closing management unit 15. DETAILED DESCRIPTION

[0012] The present application solves the technical problem in the prior art that intelligent account closing cannot be performed according to demand, resulting in low work efficiency, by providing an intelligent account closing method and system.

[0013] Embodiment 1, as Figure 1 As shown, an embodiment of the present application provides an intelligent account closing method, the method comprising:

[0014] Step S100: Establish a historical account closing database, extract data features from the historical account closing database, synchronize the data feature extraction results to the prediction model, and generate a predicted account closing time and focus on account closing anomalies.

[0015] By collecting data from multiple closing cycles in the past, such as closing date, closing time, closing status, abnormal records, etc., a historical closing database is established. Data features are extracted from the historical closing database to obtain closing time features (average time, maximum time, and minimum time required for closing), abnormal features (frequency of abnormal events, time point distribution), and user behavior features (such as manual temporary opening of the system, etc.); the extracted feature data is synchronized to the prediction model, which analyzes time series data and user behavior records to output predicted closing time and focus on closing abnormalities, providing support for closing planning and abnormality warning.

[0016] Furthermore, step S100 of the present application also includes:

[0017] Step S110: Based on the preprocessing rule base, priority identification is performed on the data feature extraction result, and a priority feature set is established. The preprocessing rule base is the first processing layer of the prediction model. Step S120: The priority feature set is used as the first input data, and the time embedding is used as the second input data, and is synchronously input to the global prediction layer, and a probability distribution prediction of the closing time point is performed to establish a probability distribution prediction result, wherein the time embedding is an enhanced input, and the global prediction layer is the second processing layer of the prediction model. Step S130: The probability distribution prediction result is synchronized to the dynamic decision layer, and a dynamic decision analysis is performed to establish a closing time, and a closing anomaly is configured to be paid attention to, wherein the dynamic decision layer is the third processing layer of the prediction model.

[0018] After the data feature extraction is completed, the predefined preprocessing rule library is used to identify the priority of the feature data. The preprocessing rule library contains a set of rules for the characteristics of closing data. These rules may be based on business logic, data correlation analysis or expert experience. By matching the rules, the data features are prioritized, the most influential features for closing prediction are screened out, and a priority feature set is established. The priority feature set reduces the interference of redundant information and provides high-quality input data for subsequent model prediction. The preprocessing rule library is used as the first processing layer of the prediction model to improve the effectiveness and pertinence of the input data. The priority feature set is used as the first input data of the prediction model, and the historical time information is processed into time embedding as the second input data of the prediction model, and is input into the global prediction layer together. As an enhanced input, time embedding can improve the model's understanding of time dependence, such as the law of periodic changes in closing time. The global prediction layer is the second processing layer of the prediction model. Through deep learning algorithms (such as LSTM, Transformer, etc.), a probability distribution prediction model for closing time points is established to generate the probability distribution results of each time point that may become the closing time. The global prediction layer is used as the second processing layer of the prediction model to improve the accuracy of the prediction and the robustness of time correlation. The probability distribution prediction results are input into the third processing layer (dynamic decision layer) of the prediction model. The dynamic decision layer combines the actual business needs with the probability distribution prediction results, executes the final decision on the closing time through dynamic analysis algorithms (such as Bayesian optimization or reinforcement learning), and generates an executable closing time plan. At the same time, combined with historical abnormal data and priority feature sets, configuration is configured to focus on closing abnormalities, marking modules, operations or users that may have problems for subsequent key monitoring.

[0019] Step S200: configure whitelist users during the closing period and establish a whitelist lookback window.

[0020] During the closing period, in order to ensure that specific users can complete key business operations when the system is closed, and to ensure system security and controllability of operations, the system configures whitelist users and establishes a whitelist backtracking window. Specifically, the whitelist users during the closing period are screened and configured. These users are usually key personnel who need to access during the closing process, such as financial administrators, senior auditors, or emergency problem handlers; the system dynamically grants special access rights during the closing period to whitelist users, allowing them to access specific modules when the system is closed, and restricts their access scope and access time according to actual needs, such as limited to the budget module or the reimbursement module, and only effective within a specific period from the start to the completion of the closing. On this basis, a whitelist backtracking window is established to record and monitor all operations of whitelist users during the closing period. The whitelist backtracking window covers the closing period and the key time periods before and after it, and records in detail the user's login time, access module, specific operations performed, and related risk assessment information. Through the collaborative mechanism of whitelist user configuration and backtracking window, the system realizes flexible release and dynamic control of key users, effectively ensuring the efficiency and security of the closing process.

[0021] Step S300: performing account closing processing based on the financial module according to the predicted account closing time, the concerned account closing anomaly, and the account closing list, and releasing the whitelist users during the account closing period. The account closing list includes a budget module, a cost module, a purchase and sales module, and a reimbursement module. The budget module, the cost module, the purchase and sales module, and the reimbursement module are linked in real time by using the financial module to complete the account closing processing.

[0022] During the closing process, the system performs intelligent closing processing based on the financial module according to the predicted closing time, the closing anomalies and the closing list, and implements flexible release for whitelist users during the closing period. Specifically, according to the predicted closing time, the identified closing anomalies and the closing list (including the budget module, cost module, inventory management module and accounting module), the financial module performs real-time linkage processing on each business module. For example, after the budget adjustment is completed in the budget module, it is synchronized to the cost module in real time; after the data is updated in the cost module, the adjustment results are passed to the inventory management module and the accounting module, so as to ensure the real-time consistency of data between modules.

[0023] Exemplarily, based on the predicted closing time, the closing process is started. During the process, the data of each module is checked, adjusted and confirmed. If any abnormality or problem is found during the closing process, it is immediately investigated and handled. After the closing is completed, the closing results are recorded, including information such as the data status of each module, the closing time, and the closing personnel.

[0024] During the closing period, the system automatically releases the operating permissions of the configured whitelist users, allowing them to access relevant modules and complete necessary operations in the closed state. The release scope of whitelist users is strictly controlled and limited to specific modules and specific operation scenarios. For example, financial administrators can enter or proofread data in the budget module and the reimbursement module, while the access permissions of other modules are restricted. At the same time, the system dynamically monitors the operating behavior of whitelist users, records their access modules, operation types and abnormal situations in real time, and ensures operational security. Finally, the budget, cost, inventory management and reimbursement modules are fully linked through the financial module, realizing data consistency verification, abnormal correction and efficient execution of the automated closed-loop closing process, ensuring that the closing work is completed smoothly on the basis of safety, accuracy and efficiency.

[0025] Step S400: monitoring the behavior of the whitelist user during the account closing period, and generating a real-time warning based on the whitelist backtracking window and the behavior monitoring result.

[0026] During the closing period, the system monitors the operation behavior of whitelist users in real time, and generates real-time warnings based on the historical records and behavior monitoring results in the whitelist backtracking window to ensure the security of the system and the standardization of operations. By comprehensively recording and analyzing all operations of whitelist users, including login time, access module, type of operation performed, and frequency of operation, potential abnormal behaviors can be identified, such as high-frequency repeated operations, access outside the scope of authority, sensitive operations beyond expectations, or operations that are obviously inconsistent with historical behavior patterns. The whitelist backtracking window is used to compare and analyze the user's historical operation records. The whitelist backtracking window records the behavior patterns of whitelist users in the past multiple closing cycles, including operation paths, access frequencies, and switching patterns between modules. For example, when a user has no similar operations in the historical records or the frequency of behavior increases abnormally, the system will mark it as a high-risk operation. Based on the real-time monitoring data and the comparison results of the backtracking window, the system generates a real-time warning.

[0027] Furthermore, step S400 of the present application also includes:

[0028] Step S410: Perform trigger analysis on the whitelist backtracking window based on the behavior monitoring results, perform behavior backtracking analysis within the whitelist backtracking window according to the trigger analysis results, and generate a basic warning result; Step S420: Perform time series trend prediction on the basic warning result, and configure the window dynamic expansion factor according to the time series trend prediction result; Step S430: Dynamically expand the whitelist backtracking window using the window dynamic expansion factor, and perform additional authentication using the dynamically expanded whitelist backtracking window to establish an additional warning result; Step S440: Generate a real-time warning based on the basic warning result and the additional warning result.

[0029] During the closing period, the system generates real-time warnings based on the whitelist backtracking window and behavior monitoring results. First, the operation behavior of the whitelist users is triggered and analyzed through real-time behavior monitoring to identify potential abnormal operations, such as access beyond the scope of authority, high-frequency operations, or non-standard access to sensitive data. Once the triggering conditions are met, the system starts the whitelist backtracking window, analyzes the historical behavior data of the user within the backtracking window, evaluates the abnormality of its behavior pattern, and generates basic warning results. Subsequently, based on the basic warning results, the time series prediction algorithm is used to analyze the future trend of the user's abnormal behavior, such as whether the abnormal behavior is likely to expand or continue. According to the trend prediction results, the system dynamically calculates the window expansion factor, adjusts the time span of the backtracking window, and captures a wider range of historical behavior data. The expanded backtracking window data is further used for additional authentication analysis to confirm the legitimacy of the operation by evaluating whether the user operation involves sensitive data or introducing multi-factor authentication, thereby generating additional warning results. Finally, the system combines the basic warning results and additional warning results, combines the abnormal behavior description, risk level and impact scope, generates real-time warnings, and pushes them to the administrator in real time, while providing processing suggestions, such as restricting user permissions or starting emergency handling procedures.

[0030] Furthermore, step S400 of the present application also includes:

[0031] Step S450: Establish multi-factor identity authentication for whitelist users, and build a trust verification channel based on the multi-factor identity authentication; Step S460: Synchronize the real-time warning to the trust verification channel and generate a multi-factor identity authentication scheme; Step S470: Use the multi-factor identity authentication scheme to perform real-time identity re-verification of whitelist users.

[0032] During the closing period, the system further strengthens the identity authentication and trust management mechanism of whitelist users to improve operational security and abnormal response capabilities. Specifically, the system establishes a multi-factor identity authentication mechanism for whitelist users, combining dynamic passwords, biometrics (such as fingerprints or face recognition) and device binding information to build a trusted trust verification channel. The trust verification channel is used to dynamically evaluate the user's operation credibility and provide support for subsequent real-time warning processing; when the system generates a real-time warning, the warning information will be synchronized to the trust verification channel, and based on the user's operation risk level and warning content, a targeted multi-factor identity authentication scheme will be dynamically generated. For example, for low-risk warnings, simple two-factor authentication can be used, while for high-risk warnings, users are required to perform high-level identity confirmation through biometric verification or device binding verification; the generated multi-factor identity authentication scheme is used to conduct real-time identity re-verification of whitelist users to ensure the legality and credibility of the current operation. If the user fails to pass the identity re-verification, the system will automatically limit his or her operation permissions and notify the administrator to take further measures. Through the synergy of multi-factor authentication and trust verification channels, the system effectively strengthens the identity management capabilities of whitelisted users, ensures system security and operational compliance during closing periods, and reduces the potential impact of abnormal behavior on system operations.

[0033] Furthermore, step S400 of the present application also includes:

[0034] Step S480: extracting whitelist user behavior features based on the behavior monitoring results, and establishing whitelist user behavior feature extraction results; Step S490: performing user group collaboration pattern analysis on the whitelist user behavior feature extraction results, and calculating the collaboration deviation; Step S4100: establishing a group behavior warning based on the collaboration deviation, and performing intelligent account closing management based on the group behavior warning.

[0035] During the closing period, the system further enhances its ability to analyze and manage the behavior of whitelisted user groups, and implements a more comprehensive abnormal warning mechanism and intelligent account closing management through collaborative analysis. Specifically, based on the behavior monitoring results, feature extraction is performed on the operation behavior of whitelist users, and core features related to user operations are mined, such as operation frequency, access module, behavior path, and interaction mode with other users, and the whitelist user behavior feature extraction results are generated; the extracted behavior feature results are analyzed for user group collaboration mode, and the collaborative operation mode of the whitelist user group is identified through cluster analysis, association rule mining and other algorithms, and the collaboration deviation of individual users is calculated to evaluate the consistency of their behavior with the group operation mode. If the collaboration deviation of a user is significantly higher than the normal range, it indicates that his behavior may be abnormal or deviate from the group collaboration goal; based on the collaboration deviation, a group behavior warning is established, and an overall risk assessment result is generated by comprehensively analyzing the abnormal situations in the user group. For example, when multiple users deviate from the group collaboration mode at the same time, the system will trigger a high-risk warning, indicating that there may be coordination problems or safety hazards; the closing management strategy is dynamically adjusted according to the group behavior warning results, such as restricting the closing authority of users with higher deviations or adjusting the scope of operation authority, thereby realizing intelligent optimization and security of the closing process. By introducing user group collaboration pattern analysis and collaboration deviation calculation, the system further improves the ability to identify potential risks in the account closing process, effectively ensuring the efficiency and security of account closing work.

[0036] Step S500: Perform intelligent account closing management according to the real-time warning.

[0037] According to the real-time warning, the system dynamically adjusts the closing process to achieve intelligent closing management, thereby improving closing efficiency and ensuring data security. Specifically, when the system generates a real-time warning, it will dynamically optimize the closing management strategy based on the risk level of the warning content, the modules involved, and the user behavior characteristics; for low-risk warnings, the system will automatically record and release related operations, and mark the corresponding users and operation modules for subsequent auditing and analysis; for medium-risk warnings, the system will suspend the current operation, prompt the user to perform multi-factor identity authentication or additional operation confirmation, and restore the operation authority after verification; for high-risk warnings, the system will immediately suspend the operation authority of the relevant users, lock the abnormal module, and notify the administrator to intervene.

[0038] Furthermore, step S500 of the present application also includes:

[0039] Step S510: Perform business impact fitting according to the real-time warning and establish a safe business range interval; Step S520: After freezing and isolating the safe business range interval, disable the corresponding whitelist user rights and issue a warning.

[0040] By analyzing the whitelist user behaviors, operation modules and their correlation with other modules and data involved in the early warning, the potential impact of abnormal behaviors on the closing business is evaluated, and the corresponding safe business scope interval is established. The safe business scope interval defines the safe modules, data ranges and functional boundaries that can continue to operate under abnormal circumstances to ensure that abnormal behaviors will not affect the normal operation of the overall closing process.

[0041] Freeze and isolate the identified security business scope. The main measures of freeze and isolate include: temporarily locking the sensitive operation permissions of the module where the abnormal user is located, such as data writing or transmission operations, to prevent the abnormal behavior from expanding the impact; data isolation processing for other modules involved in the security scope to ensure that the modules or data operated by the abnormal user will not further affect the associated modules; stop the corresponding permissions of the whitelist users to terminate their abnormal operation behavior. At the same time, an early warning report is generated based on the freeze and isolate results, and the affected modules, security business scope intervals, abnormal users and specific operation risk levels in the closing process are sent to the system administrator through real-time notification.

[0042] Furthermore, step S500 of the present application also includes:

[0043] Step S530: record the business flow data during the closing period and establish a flow data set; Step S540: obtain user feedback during the closing period, and establish a self-optimizing data set based on the user feedback and the flow data set; Step S550: perform intelligent closing optimization iteration based on the self-optimizing data set.

[0044] During the closing process, the business flow data between modules is recorded in real time, including detailed information such as the task execution sequence, data transmission process, processing delay and operation results of the budget module, cost module, inventory module and accounting module, so as to establish a flow data set that fully reflects the dynamic interaction of the closing process. User operation feedback during the closing process is obtained through the user interaction interface or feedback survey, including the evaluation of the convenience of the operation process, the satisfaction with the system response speed, and the questions and suggestions raised by the user. Combining the flow data set with user feedback information, the system conducts correlation analysis, extracts the bottleneck points or optimization space in the closing process, and constructs a self-optimizing data set. Through data labeling, classification and feature extraction, it provides a basis for the optimization and iteration of the closing process. Based on the self-optimizing data set, the closing process is iteratively optimized through intelligent algorithms, including task scheduling optimization to reduce task conflicts between modules, operation process optimization to simplify user interaction, data flow optimization to reduce transmission delay, and exception handling optimization to enhance the ability to respond to risk scenarios. Through this closed-loop optimization management, the system can continuously improve the efficiency, stability and user experience of the closing process, dynamically adapt to different business scenarios and user needs, and ultimately achieve efficient, secure and intelligent management of the closing process.

[0045] In summary, the embodiments of the present application have at least the following technical effects:

[0046] First, a historical account closing database is established. After extracting data features from the historical account closing database, the data feature extraction results are synchronized to the prediction model to generate the predicted account closing time and the account closing anomalies to be paid attention to. Next, whitelist users during the account closing period are configured, and a whitelist backtracking window is established. Then, account closing processing based on the financial module is performed according to the predicted account closing time, the account closing anomalies to be paid attention to, and the account closing list, and whitelist users are released during the account closing period. The account closing list includes the budget module, the cost module, the purchase and sales module, and the reimbursement module. The budget module, the cost module, the purchase and sales module, and the reimbursement module are linked in real time by the financial module to complete the account closing processing. Furthermore, the behavior of whitelist users during the account closing period is monitored, and real-time warnings are generated based on the whitelist backtracking window and the behavior monitoring results. Finally, intelligent account closing management is performed based on the real-time warning. The technical problem that the existing technology cannot perform intelligent account closing according to demand, resulting in low work efficiency, is solved. By configuring whitelist users for intelligent account closing management, the technical effect of improving work efficiency and reducing business risks is achieved.

[0047] Embodiment 2 is based on the same inventive concept as the intelligent account closing method in the above embodiment. Figure 2 As shown, the present application provides an intelligent account closing system, the system comprising:

[0048] The database establishment unit 11 is used to establish a historical account closing database, extract data features from the historical account closing database, and synchronize the data feature extraction results to the prediction model to generate a predicted account closing time and a concerned account closing anomaly; the whitelist configuration unit 12 is used to configure whitelist users during the account closing period and establish a whitelist backtracking window; the account closing processing unit 13 is used to perform account closing processing based on the financial module according to the predicted account closing time, the concerned account closing anomaly, and the account closing list, and release the whitelist users during the account closing period. The account closing list includes a budget module, a cost module, a purchase and sales inventory module, and a reimbursement module. The budget module, the cost module, the purchase and sales inventory module, and the reimbursement module are linked in real time by the financial module to complete the account closing processing; the behavior monitoring unit 14 is used to monitor the behavior of the whitelist users during the account closing period, and generate a real-time warning based on the whitelist backtracking window and the behavior monitoring results; the account closing management unit 15 is used to perform intelligent account closing management according to the real-time warning.

[0049] Furthermore, the database establishing unit 11 is used to execute the following method:

[0050] Based on the preprocessing rule base, the priority of the data feature extraction result is identified, and a priority feature set is established. The preprocessing rule base is the first processing layer of the prediction model; the priority feature set is used as the first input data, and the time embedding is used as the second input data, which are synchronously input to the global prediction layer, and the probability distribution prediction of the closing time point is performed to establish the probability distribution prediction result, wherein the time embedding is the enhanced input, and the global prediction layer is the second processing layer of the prediction model; the probability distribution prediction result is synchronized to the dynamic decision layer, and dynamic decision analysis is performed to establish the closing time, and the attention to the closing anomaly is configured, wherein the dynamic decision layer is the third processing layer of the prediction model.

[0051] Furthermore, the behavior monitoring unit 14 is used to perform the following method:

[0052] Based on the behavior monitoring results, a trigger analysis of the whitelist backtracking window is performed, and according to the trigger analysis results, a behavior backtracking analysis is performed within the whitelist backtracking window to generate a basic warning result; a time series trend prediction is performed on the basic warning result, and a window dynamic expansion factor is configured according to the time series trend prediction result; the whitelist backtracking window is dynamically expanded using the window dynamic expansion factor, and additional authentication is performed using the dynamically expanded whitelist backtracking window to establish an additional warning result; a real-time warning is generated based on the basic warning result and the additional warning result.

[0053] Furthermore, the behavior monitoring unit 14 is used to perform the following method:

[0054] Establish multi-factor identity authentication for whitelist users, and build a trust verification channel based on the multi-factor identity authentication; synchronize the real-time warning to the trust verification channel and generate a multi-factor identity authentication scheme; use the multi-factor identity authentication scheme to perform real-time identity re-verification of whitelist users.

[0055] Furthermore, the behavior monitoring unit 14 is used to perform the following method:

[0056] Based on the behavior monitoring results, whitelist user behavior feature extraction is performed to establish whitelist user behavior feature extraction results; user group collaboration pattern analysis is performed on the whitelist user behavior feature extraction results to calculate the collaboration deviation; group behavior warning is established based on the collaboration deviation, and intelligent account closing management is performed based on the group behavior warning.

[0057] Furthermore, the account closing management unit 15 is used to execute the following method:

[0058] Business impact fitting is performed based on the real-time warning to establish a safe business range; after freezing and isolating the safe business range, the corresponding whitelist user rights are disabled and a warning is issued.

[0059] Furthermore, the account closing management unit 15 is used to execute the following method:

[0060] The business flow data during the closing period is recorded and a flow data set is established; user feedback during the closing period is obtained, and a self-optimization data set is established according to the user feedback and the flow data set; and intelligent closing optimization iteration is performed according to the self-optimization data set.

[0061] Through the above detailed description of an intelligent account closing method in this specification, those skilled in the art can clearly understand an intelligent account closing system in this embodiment. As for the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the description of the method part.

[0062] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An intelligent account closing method, characterized in that: The method comprises: Establish a historical account closing database, extract data features from the historical account closing database, synchronize the data feature extraction results to the prediction model, generate a predicted account closing time and pay attention to account closing anomalies; Configure whitelist users during the closing period and establish a whitelist lookback window; Performing account closing processing based on the financial module according to the predicted account closing time, the concerned account closing anomaly, and the account closing list, and releasing the whitelisted users during the account closing period, the account closing list including the budget module, the cost module, the purchase and sales module, and the reimbursement module, and completing the account closing processing after the budget module, the cost module, the purchase and sales module, and the reimbursement module are linked in real time by the financial module; Monitor the behavior of the whitelist users during the account closing period, and generate real-time warnings based on the whitelist backtracking window and behavior monitoring results; Intelligent account closing management is performed based on the real-time warning.

2. The intelligent account closing method according to claim 1, characterized in that: The data feature extraction results are synchronized to the prediction model to generate the predicted closing time and focus on closing anomalies, including: Priority identification of the data feature extraction results is performed based on a preprocessing rule base to establish a priority feature set, wherein the preprocessing rule base is the first processing layer of the prediction model; The priority feature set is used as the first input data, and the time embedding is used as the second input data, which are synchronously input to the global prediction layer, and the probability distribution prediction of the closing time point is performed to establish the probability distribution prediction result, wherein the time embedding is the enhanced input, and the global prediction layer is the second processing layer of the prediction model; The probability distribution prediction result is synchronized to the dynamic decision layer, dynamic decision analysis is performed, closing time is established, and closing anomalies are configured, wherein the dynamic decision layer is the third processing layer of the prediction model.

3. The intelligent account closing method according to claim 1, characterized in that: The generating of a real-time warning based on the whitelist backtracking window and the behavior monitoring result includes: Perform trigger analysis on the whitelist backtracking window based on the behavior monitoring result, perform behavior backtracking analysis within the whitelist backtracking window according to the trigger analysis result, and generate a basic warning result; Performing time series trend forecasting on the basic warning results, and configuring a window dynamic expansion factor according to the time series trend forecasting results; Dynamically expand the whitelist backtracking window using the window dynamic expansion factor, and use the dynamically expanded whitelist backtracking window to perform additional authentication and establish additional warning results; A real-time warning is generated according to the basic warning result and the additional warning result.

4. The intelligent account closing method according to claim 1, characterized in that: After performing intelligent account closing management according to the real-time warning, the method further includes: Establishing multi-factor authentication for whitelisted users and building a trust verification channel based on the multi-factor authentication; Synchronizing the real-time warning to the trust verification channel to generate a multi-factor identity authentication scheme; The multi-factor identity authentication scheme is used to perform real-time identity verification of whitelisted users.

5. The intelligent account closing method according to claim 1, characterized in that: The method further comprises: Based on the behavior monitoring results, whitelist user behavior feature extraction is performed to establish whitelist user behavior feature extraction results; Performing user group collaboration mode analysis on the whitelist user behavior feature extraction results and calculating the collaboration deviation; A group behavior warning is established according to the collaboration deviation, and intelligent account closing management is performed according to the group behavior warning.

6. The intelligent account closing method according to claim 1, characterized in that: The intelligent account closing management according to the real-time warning includes: Perform business impact fitting based on the real-time warning and establish a safe business range; After freezing and isolating the security business scope, the corresponding whitelist user permissions are disabled and an early warning is issued.

7. The intelligent account closing method according to claim 1, characterized in that: After performing intelligent account closing management according to the real-time warning, the method further includes: Record business flow data during the closing period and establish a flow data set; Obtain user feedback during the closing period, and establish a self-optimizing data set according to the user feedback and the circulating data set; Intelligent account closing optimization iteration is performed according to the self-optimization data set.

8. An intelligent account closing system, characterized in that: A system for implementing an intelligent account closing method according to any one of claims 1 to 7, comprising: Database establishment unit: establishes a historical account closing database, extracts data features from the historical account closing database, synchronizes the data feature extraction results to the prediction model, generates a predicted account closing time and focuses on account closing anomalies; Whitelist configuration unit: configure whitelist users during the closing period and establish a whitelist lookback window; The account closing processing unit performs account closing processing based on the financial module according to the predicted account closing time, the concerned account closing anomaly and the account closing list, and releases the whitelist users during the account closing period. The account closing list includes a budget module, a cost module, a purchase and sales module and a reimbursement module. The budget module, the cost module, the purchase and sales module and the reimbursement module are linked in real time by the financial module to complete the account closing processing; Behavior monitoring unit: monitors the behavior of the whitelist users during the account closing period, and generates real-time warnings based on the whitelist backtracking window and behavior monitoring results; Account closing management unit: performs intelligent account closing management according to the real-time warning.

Images