Auditing process processing method and system suitable for financial management

Abstract

The invention relates to the technical field of financial management, and discloses an audit process processing method and system suitable for financial management, and the method comprises the steps: extracting an audit event record and process current state data according to a business operation event and a process instance state; and performing association mapping on the audit event record and the current dynamic node to obtain an instance audit mark node. And positioning flow audit constraints corresponding to the instance audit mark node based on the flow definition file, wherein the flow audit constraints comprise a to-be-verified field name, an operator and a verification reference value. And performing logical operation on the actual numerical value corresponding to the to-be-verified field name in the audit event record according to the operator and the verification reference value to generate a risk state identifier. And arranging the task nodes according to the starting and ending timestamps of the historical node flow record and the logic connection relationship to obtain a flow path. Attaching the risk state identifier to a corresponding node of the flow path, and generating a dynamic audit process view; according to the invention, the accuracy of audit process processing suitable for financial management can be improved.

Description

Technical Field

[0001] This invention relates to the field of financial management technology, and in particular to an audit process processing method and system applicable to financial management. Background Technology

[0002] In the current field of financial audit management, traditional audit process methods largely rely on post-event static data inspection and manual experience judgment. These methods typically involve auditors exporting data from financial systems or log files after the business process has completely ended, and then manually comparing or using simple queries to identify anomalies. This approach cannot correlate audit events with the dynamic running status of process instances in real time, resulting in a disconnect between audit event records and the current node in the process. Auditors find it difficult to accurately determine at which specific stage of the process a particular operation was triggered. Furthermore, traditional audit rule execution often uses fixed thresholds or single verification conditions, lacking the ability to dynamically resolve node-level constraints in process definition files. When business rules change, manual modifications to code or configuration files are required, leading to slow response times and a high risk of errors. Because the risk status cannot be obtained in real time during process execution, audit work often lags behind business operations, failing to promptly block or warn of potential risky behaviors, thus posing significant hidden dangers to financial management.

[0003] Another major drawback of existing technologies is their inability to generate a visual, dynamic audit process view. Traditional audit results are typically presented in tabular or text reports, lacking an intuitive display of the process flow path. Auditors struggle to sift through large amounts of log records to trace the historical flow sequence of process instances, and cannot link risk status with specific node locations. When analyzing the processing efficiency or idle waiting time of a particular node, traditional methods often require manual calculation of timestamp differences, which is not only inefficient but also prone to misjudgments due to calculation errors. Due to the lack of a systematic means of evaluating processing efficiency, managers cannot accurately identify bottleneck nodes in the process, nor can they quantify penalties for operations deviating from standard durations, resulting in a lack of data support for process optimization decisions. Overall, existing technologies have significant shortcomings in terms of real-time performance, accuracy, visualization, and efficiency assessment in auditing, failing to meet the demands of modern financial management for dynamic and refined audit processes. Summary of the Invention

[0004] This invention provides an audit process processing method and system suitable for financial management, the main purpose of which is to solve the problem of low accuracy in the audit process processing applicable to financial management.

[0005] To achieve the above objectives, the present invention provides an audit process handling method applicable to financial management, comprising: Based on the business operation events and process instance status of the target process, extract the audit event records and current process status data of the target process; The audit event record is associated with the current dynamic node of the current state data of the process to obtain the instance audit marker node of the target process; Based on the process definition file corresponding to the process instance status, the process audit constraint corresponding to the instance audit marker node is located. The process audit constraint includes the field name to be verified, the operator, and the verification benchmark value. The actual value corresponding to the field name to be verified in the audit event record is logically operated on with the verification benchmark value according to the operator, and a risk status identifier of the target process is generated based on the operation result; Based on the start and end timestamps of the historical node flow records in the process instance status, and combined with the logical connection relationships between task nodes in the historical node flow records, the task nodes are arranged to obtain the flow path of the target process. The risk status identifier is attached to the corresponding node of the flow path to obtain a dynamic audit process view of the target process.

[0006] In a preferred embodiment, the step of extracting the audit event records and current process status data of the target process based on the business operation events and process instance status of the target process includes: Data change records are captured from the database transaction log of the target process, and the operation type identifier, operation timestamp, operator account name and financial data field value in the data change records are combined into the business operation event of the target process; Traverse the process instance states in the target process, extract the current node identifier, node trigger timestamp, associated business data index, and node traversal trajectory of the business instance in the process instance state of the target process, and obtain the current process state data of the target process; The operation timestamp of the business operation event is compared with the node trigger timestamp. If the operation timestamp is later than the node trigger timestamp, the business operation event is recorded as an audit event of the target process. If the operation timestamp is earlier than or equal to the node trigger timestamp, the node with a termination timestamp later than the operation timestamp is searched from the node's experience trajectory, and the business operation event is assigned to the audit event record of the node.

[0007] In a preferred embodiment, the step of associating and mapping the audit event record with the current dynamic node of the current state data of the process to obtain the instance audit marker node of the target process includes: Align the timestamp of the audit event record with the time point of the status change of each dynamic node in the current status data of the process, and filter out the target dynamic node that matches the time of the audit event record. The event type identifier of the audit event record is compared with the node attribute information of the target dynamic node. Audit event records with matching characteristics are anchored to the target dynamic node to generate an instance audit marker node for the target process.

[0008] In a preferred embodiment, the process audit constraint corresponding to the instance audit marker node is located based on the process definition file corresponding to the process instance state. The process audit constraint includes the field name to be verified, the operator, and the verification benchmark value, including: Based on the process instance status, the process definition identifier reads the corresponding process definition file, and matches the corresponding node configuration information segment in the process definition file according to the node identifier of the instance audit mark node. Scan the condition expression field in the node configuration information segment to obtain the original constraint text of the target process; The original constraint text is parsed to identify the verification field names, operators, and verification benchmark values ​​in the original constraint text, thereby obtaining the process audit constraints of the target process.

[0009] In a preferred embodiment, parsing the original constraint text to identify the validation field names, operators, and validation benchmark values ​​in the original constraint text, and obtaining the process audit constraints of the target process, includes: Traverse the character sequence of the original constraint text and extract the operators in the character sequence; Using the position of the first relational operation character in the original constraint text as the dividing point, the original constraint text is divided into a left substring and a right substring; Starting from the end of the left substring, scan backwards and use the alphanumeric string adjacent to the relational operation character as the verification field name of the target process; Starting from the first character of the right substring, scan sequentially backwards, and use the consecutive numerical character sequence adjacent to the relational operation character as the verification benchmark value of the target process.

[0010] In a preferred embodiment, the step of performing logical operations on the actual value corresponding to the field name to be verified in the audit event record and the verification benchmark value according to the operator, and generating a risk status identifier for the target process based on the operation result, includes: Extract the actual value that matches the field name to be verified from the audit event record, and compare the actual value with the amount threshold and approval time limit threshold in the verification benchmark value according to the operator to obtain the comparison result value of the target process; If the comparison result is true, the risk level code of the target process is extracted from the process audit constraint, and the current time timestamp of the target process is obtained based on the marker time of the risk level code. The risk level code, the current time timestamp, and the node identifier of the instance audit marker node are combined to form the risk status identifier of the target process. If the comparison result is false, a null value risk status identifier is generated and written into the risk status field of the instance audit marker node.

[0011] In a preferred embodiment, the step of arranging the task nodes to obtain the flow path of the target process based on the start and end timestamps of the historical node flow records in the process instance state, and in conjunction with the logical connection relationships between task nodes in the historical node flow records, includes: Extract the historical node flow record of the target process from the state of the process instance. The historical node flow record includes node identifier, node type, start timestamp and end timestamp. Read the connection arc definition between task nodes from the process definition file corresponding to the process instance state. The connection arc definition includes a source node identifier and a target node identifier. Taking the task node with the earliest start timestamp as the path start node, and according to the binding relationship between the source node identifier and the target node identifier in the connection arc definition, the historical node flow record is traversed to obtain the node sequence of the target process; During the traversal, for task nodes with parallel branches, the child nodes associated with the task node in the process definition file are located based on the parallel gateway identifier of the node type field value in the task node. The child nodes are arranged in the same level according to the order of the start timestamps. For task nodes whose end timestamp is later than the start timestamp of the successor node, the logical relationship defined by the connection arc is used as the connection basis, and the task node is still arranged first and the successor node is arranged second. The node sequence is combined according to time order and logical relationship, and the status information of the processing efficiency of the task node is written into the node data of the corresponding node in the node sequence during the combination to obtain the flow path of the target process.

[0012] In a preferred embodiment, the formula for calculating the processing efficiency includes: in, For the first The processing efficiency coefficient of each node. For node indexing, Based on basic efficiency weights, For the first Standard processing time configuration parameters for each node For the first The actual time each node is occupied To deviate from the penalty weight, It is a natural constant. To await the penalty weight, It is the natural logarithm function. For the first The idle time of each node.

[0013] In a preferred embodiment, attaching the risk status identifier to the corresponding node of the flow path to obtain a dynamic audit process view of the target process includes: Based on the value of the risk status identifier, the corresponding nodes of the flow path are marked to obtain the initial audit process view of the target process; The trigger timestamp of the risk status identifier is converted into a standard date and time format and written into the remarks field of the corresponding node in the initial audit process view to obtain the dynamic audit process view of the target process.

[0014] To address the aforementioned problems, the present invention also provides an audit process processing system suitable for financial management, the system comprising: The data configuration module extracts the audit event records and current process status data of the target process based on the business operation events and process instance status of the target process. The instance audit tagging module associates and maps the audit event record with the current dynamic node of the current state data of the process to obtain the instance audit tagging node of the target process; The process audit constraint module locates the process audit constraint corresponding to the instance audit marker node based on the process definition file corresponding to the process instance status. The process audit constraint includes the field name to be verified, the operator, and the verification benchmark value. The risk status identification module performs logical operations on the actual value corresponding to the field name to be verified in the audit event record and the verification benchmark value according to the operator, and generates the risk status identification of the target process based on the operation result. The flow path module arranges the task nodes according to the start and end timestamps of the historical node flow records in the process instance status, and in combination with the logical connection relationship between the task nodes in the historical node flow records, to obtain the flow path of the target process. The dynamic audit process view module attaches the risk status identifier to the corresponding node of the flow path to obtain a dynamic audit process view of the target process.

[0015] Compared with the prior art, the present invention has the following beneficial effects:

[0016] 1. This technical solution achieves high-precision correlation between audit events and process nodes by capturing data change records in real time from the database transaction log and comparing them with the dynamic running status of process instances through time-series comparison and feature anchoring. This dynamic correlation mechanism ensures that every business operation can be accurately assigned to a specific process stage, so that audit analysis no longer relies on manual retrospective tracing, but can obtain complete operation records of each node during process execution. Simultaneously, the solution automatically extracts verification field names, operators, and verification benchmark values ​​by dynamically parsing the node-level constraint text in the process definition file, achieving deep binding between audit rules and process design. When the business process is adjusted, only the constraints in the definition file need to be modified to update the audit logic synchronously, significantly improving the flexibility and responsiveness of rule configuration and ensuring that audit verification is always consistent with the latest business requirements.

[0017] 2. This technical solution generates a dynamic audit process view rich in multi-source information by integrating time, logic, and efficiency dimensions. In reconstructing the process flow path, the solution not only accurately restores the execution order and parallel structure of nodes based on timestamps and connection arcs, but also embeds processing efficiency status information into the data of each node, enabling the process view to intuitively display the efficiency performance of each node. The value status of risk status identifiers is further transformed into visual node markers, and the trigger time is converted into a standard format and written into the remarks field. The final dynamic audit process view fully presents the node sequence, logical relationships, risk levels, and processing efficiency. This multi-dimensional visualization output provides managers with comprehensive process insights, enabling precise data support for bottleneck node identification, risk positioning, and resource allocation optimization, significantly improving the refinement level of audit process management. Attached Figure Description

[0018] Figure 1 A flowchart illustrating an audit process processing method applicable to financial management, provided as an embodiment of the present invention;

[0019] Figure 2A functional module diagram of an audit process processing system suitable for financial management, provided as an embodiment of the present invention;

[0020] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0021] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0022] This application provides an audit process processing method suitable for financial management. The executing entity of this audit process processing method for financial management includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application embodiment: a server, a terminal, etc. In other words, the audit process processing method for financial management can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0023] Reference Figure 1 The diagram shown is a flowchart illustrating an audit process processing method suitable for financial management according to an embodiment of the present invention. In this embodiment, the audit process processing method suitable for financial management includes: In this embodiment of the invention, the step of extracting the audit event records and current process status data of the target process based on the business operation events and process instance status of the target process is specifically used for: Data change records are captured from the database transaction log of the target process, and the operation type identifier, operation timestamp, operator account name and financial data field value in the data change records are combined into the business operation event of the target process; Traverse the process instance states in the target process, extract the current node identifier, node trigger timestamp, associated business data index, and node traversal trajectory of the business instance in the process instance state of the target process, and obtain the current process state data of the target process; The operation timestamp of the business operation event is compared with the node trigger timestamp. If the operation timestamp is later than the node trigger timestamp, the business operation event is recorded as an audit event of the target process. If the operation timestamp is earlier than or equal to the node trigger timestamp, the node with a termination timestamp later than the operation timestamp is searched from the node's experience trajectory, and the business operation event is assigned to the audit event record of the node.

[0024] Specifically, data change records are captured from the database transaction log of the target process. The transaction log output is read in real time by deploying the database log capture component. Each data change record is parsed to obtain the operation type identifier, operation timestamp, operator account name, and financial data field value.

[0025] Specifically, it iterates through the process instance states in the target process, accesses the instance state data table of the process engine, reads the record of each process instance line by line, extracts the node identifier of the current process instance from each record, extracts the node trigger timestamp when the node is triggered, and extracts the business data index associated with the process instance.

[0026] Specifically, if the operation timestamp is later than the node trigger timestamp, the business operation event is directly marked as an audit event record of the target process. If the operation timestamp is earlier than or equal to the node trigger timestamp, the records of all historical nodes are obtained from the node experience trajectory of the current state data of the process.

[0027] Furthermore, this information is encapsulated and combined according to the data structure of business operation events to form a complete business operation event. This process is continuously captured until all relevant time ranges of the target process are covered, ultimately resulting in a complete set of business operation events for the target process.

[0028] Furthermore, the node experience trajectory field, which records the historical flow information of the process instance, is extracted, and the node sequence data stored in the field is parsed into a structured node experience trajectory. The extracted current node identifier, node trigger timestamp, associated business data index, and node experience trajectory are then merged and packaged to form the current status data of the target process.

[0029] Furthermore, each historical node record contains the termination timestamp of that node. According to the order of the historical nodes in the node experience trajectory, the operation timestamp is compared with the termination timestamp of each historical node in turn. When the first historical node with a termination timestamp later than the operation timestamp is found, the historical node is identified as the hit node, and the business operation event is assigned to the audit event record of this hit node. In this way, the audit event record is associated with the correct process node, and the audit event record belonging to each node or the current node is obtained.

[0030] In summary, this approach avoids the risks of missing or tampered information that may arise from relying on upper-level logs of business systems in traditional auditing. By directly obtaining raw data from the underlying transaction logs, it ensures the integrity, authenticity, and non-repudiation of audit data. Furthermore, by combining scattered log records into business operation events in a unified format, it provides standardized data input for subsequent correlation with process status, resolving the issues of scattered audit data sources and inconsistent formats in existing technologies.

[0031] In summary, this approach not only acquires the current node location of the process instance but also retains the node's trigger time, associated business data indexes, and a complete node flow trajectory. This provides comprehensive status information support for accurately matching business operation events to specific process nodes. Compared to traditional techniques that only record the final state, this complete data extraction method, which includes historical trajectories and current states, enables auditors to trace the entire lifecycle of the process instance, laying a solid foundation for subsequent time-series comparisons and event attribution.

[0032] In summary, this approach can accurately distinguish the timeline between an operation and a node's state change: when an operation occurs after a node is triggered, it is directly attributed to the current node; when an operation occurs before or at the same time as a node trigger, it traces back through historical nodes to find the matching termination timestamp. This dual attribution logic solves the problem of accurately determining which process stage an operation event belongs to in traditional technologies, ensuring that every audit event record can find its true process node. This provides a precise data source for subsequent node-based risk verification, significantly improving the accuracy and traceability of audits.

[0033] In this embodiment of the invention, the step of associating and mapping the audit event record with the current dynamic node of the current state data of the process to obtain the instance audit marker node of the target process is specifically used for: Align the timestamp of the audit event record with the time point of the status change of each dynamic node in the current status data of the process, and filter out the target dynamic node that matches the time of the audit event record. The event type identifier of the audit event record is compared with the node attribute information of the target dynamic node. Audit event records with matching characteristics are anchored to the target dynamic node to generate an instance audit marker node for the target process.

[0034] Specifically, first, all dynamic nodes and their corresponding status change timestamps are extracted from the current status data of the process. These dynamic nodes are then arranged in ascending order of their status change timestamps to form a timeline sequence. Next, each audit event record is traversed, and the event occurrence timestamp of the record is extracted. The first dynamic node in the timeline sequence whose status change timetamp is later than the event occurrence timestamp is searched, and this node is used as a candidate node for time matching.

[0035] Specifically, the value of the event type identifier field is extracted from the audit event record, and the value of the node type field is extracted from the node attribute information of the target dynamic node. The values ​​of these two fields are compared by string matching. If the two strings are completely identical, the characteristics are determined to be consistent. At this time, the audit event record is anchored to the target dynamic node.

[0036] Furthermore, the time interval between the node and the event timestamp is recorded simultaneously. Then, the last dynamic node whose status change time is earlier than the event timestamp is searched. The time interval is calculated again, and the two time intervals are compared. The dynamic node with the smaller time interval is selected as the target dynamic node that matches the audit event record time. If there is only one node, the node is directly selected. In this way, a unique target dynamic node is determined for each audit event record.

[0037] Furthermore, the anchoring operation involves adding the reference pointer of the audit event record to the associated event list of the target dynamic node in the current state data of the process, and setting a flag field in the data structure of the target dynamic node to identify that the node has become an instance audit flag node. At the same time, key information in the audit event record, such as the event timestamp and the operator's account name, is copied to the extended attributes of the target dynamic node. Finally, an instance audit flag node for the target process is generated. This instance audit flag node contains all the information of the original dynamic node and all the information of the associated audit event record. If the feature comparison is inconsistent, the matching result is discarded and an exception log is recorded.

[0038] In summary, this approach eliminates matching errors caused by system delays or time asynchrony, ensuring that each audit event can find the dynamic node state change point closest to its occurrence time. Compared to traditional methods that only perform simple time sequence judgments, this two-way time comparison and time interval minimization filtering mechanism significantly improves the timing accuracy of the association between audit events and process nodes, providing a reliable time matching basis for subsequent node-level risk verification.

[0039] In summary, this ensures that only audit events matching the node's business attributes are anchored, avoiding potential misattribution based solely on time matching, such as mistakenly associating operations of mismatched types to irrelevant nodes. The generated instance audit marker node integrates the status information of dynamic nodes and associated audit event information, forming a comprehensive data entity that includes both process execution status and audit operation records. This entity serves as the core unit for subsequent risk verification and view construction, enabling the risk status of each node to be independently assessed based on its associated audit events. This significantly improves the granularity and accuracy of audit analysis, while providing structured data support for the generation of dynamic audit process views.

[0040] In this embodiment of the invention, when the process audit constraint corresponding to the instance audit marker node is located based on the process definition file corresponding to the process instance state, and the process audit constraint includes the field name to be verified, operator, and verification benchmark value, it is specifically used for: Based on the process instance status, the process definition identifier reads the corresponding process definition file, and matches the corresponding node configuration information segment in the process definition file according to the node identifier of the instance audit mark node. Scan the condition expression field in the node configuration information segment to obtain the original constraint text of the target process; The original constraint text is parsed to identify the verification field names, operators, and verification benchmark values ​​in the original constraint text, thereby obtaining the process audit constraints of the target process.

[0041] Specifically, the process definition identifier is extracted from the process instance state, and the corresponding process definition file is found in the process definition repository using the identifier. The process definition file is stored in a predefined format, and the file is loaded and parsed into a set of node configurations in memory.

[0042] Specifically, locate the key-value pair or label named "Conditional Expression Field" in the obtained node configuration information segment. This field usually appears as a fixed label name or expression field in the process definition file.

[0043] Specifically, when parsing the original constraint text, the text string is scanned character by character from left to right. First, the first relational operator appearing in the text is identified. Relational operators include symbols such as greater than, less than, equal to, greater than or equal to, less than or equal to, and not equal to. The start and end positions of the operator are recorded. Then, using the operator as a dividing point, the substring to the left of the operator is extracted. Starting from the end of the substring, the characters are scanned backwards, skipping whitespace characters until the first non-whitespace character is encountered. The scan continues backwards from that position until a whitespace character or the beginning of the string is encountered. The continuous character sequence scanned is used as the validation field name.

[0044] Furthermore, each node configuration element in the set is then traversed, and the identifier field of each node configuration element is compared with the node identifier of the instance audit mark node. When a node configuration element with a completely matching identifier is found, the entire element is extracted as a node configuration information segment, which contains all the configuration attributes defined by the node in the process definition.

[0045] Further, the value of the field is read. If the value of the field is contained within a child tag, the text within the child tag is extracted. If the value of the field exists in the form of an attribute, the attribute value is extracted directly. After removing leading and trailing whitespace characters and comment symbols from the extracted text, the clean original constraint text is obtained. The original constraint text is a string that contains the conditional expression required for auditing and validating the node.

[0046] Next, the substring to the right of the operator is extracted, and the substring is scanned character by character from the beginning to the end, skipping the leading whitespace character, and continuing to scan until the first non-numeric character or the end of the string is encountered. The scanned continuous sequence of numeric characters is used as the verification benchmark value. If the verification benchmark value contains a decimal point, the numeric and decimal point are scanned continuously. Finally, the extracted verification field name, operator and verification benchmark value are combined into the process audit constraint.

[0047] In summary, this approach eliminates the hard-coded nature of audit constraints for each node within the program. Instead, they are retrieved in real-time from the process definition file. When business processes are adjusted or audit rules change, only the configuration information segment of the corresponding node in the process definition file needs modification, without altering the audit processing system's code. This significantly improves the flexibility and responsiveness of audit rule maintenance. Furthermore, precise matching based on node identifiers ensures that each instance audit tag node receives the audit rules specifically designed for it, avoiding rule mismatches or omissions.

[0048] In summary, extracting the original, unmodified constraint text directly from the process definition file preserves the complete semantics and formatting of the rules at the time of writing, avoiding information loss or distortion that may occur due to intermediate conversion or parsing. As the sole input for subsequent parsing, the original constraint text ensures that all audit conditions defined during the process design phase are fully transmitted to the audit processing system, providing a reliable textual foundation for subsequent accurate parsing and ensuring the integrity and accuracy of the audit rules.

[0049] In summary, the system can automatically extract key validation elements from complex text expressions, enabling it to clearly identify which fields to check, which comparison methods to use, and which benchmark values ​​to refer to for judgment. Compared to traditional techniques that require manual reading and understanding of text rules before configuring validation logic, this automatic parsing mechanism significantly reduces the possibility of human intervention and errors. It achieves an automated closed loop from audit rule definition to execution, providing directly computable rule inputs for subsequent risk status identification generation, and significantly improving the automation level of the entire audit process.

[0050] In this embodiment of the invention, the step of parsing the original constraint text to identify the verification field names, operators, and verification benchmark values ​​in the original constraint text, and obtaining the process audit constraints of the target process, is specifically used for: Traverse the character sequence of the original constraint text and extract the operators in the character sequence; Using the position of the first relational operation character in the original constraint text as the dividing point, the original constraint text is divided into a left substring and a right substring; Starting from the end of the left substring, scan backwards and use the alphanumeric string adjacent to the relational operation character as the verification field name of the target process; Starting from the first character of the right substring, scan sequentially backwards, and use the consecutive numerical character sequence adjacent to the relational operation character as the verification benchmark value of the target process.

[0051] Specifically, starting from the first character of the original constraint text, each character is read sequentially. Each character is compared with a predefined set of operator characters, which includes symbols such as greater than, less than, equal to, exclamation mark, combinations of greater than and equal to, combinations of less than and equal to, and combinations of exclamation mark and equal to. When a character is found to belong to this set, the system continues to check whether the next character can form a two-character operator with that character.

[0052] Specifically, after obtaining the list of operators, the first operator in the list is taken out, and the starting position of the operator in the original constraint text is obtained. Using the starting position as the dividing point, the substring from the beginning of the original constraint text to the beginning position is extracted as the left substring, and the substring from the beginning position to the end of the original constraint text is extracted as the right substring.

[0053] Specifically, take the last character of the left substring as the current position, and move backward character by character from that position. At each character, check whether the character is a letter or a number. If it is a letter or a number, continue moving forward. If a space or a non-alphanumeric character is encountered, stop moving and extract the continuous character sequence from the stop position to the end of the left substring.

[0054] Specifically, take the first character of the right substring as the current position, and move forward character by character from that position. At each character, check if the character is a number or a decimal point. If it is a number or a decimal point, continue moving forward. If a space or a non-number or non-decimal point character is encountered, stop moving. Extract the continuous character sequence from the beginning of the right substring to the stop position.

[0055] Furthermore, for example, the greater than sign and the equal sign form the greater than or equal to sign. If they form a two-character operator, then both characters are extracted as one operator. If the next character cannot form a two-character operator, then only the current single character is extracted as the operator. The start and end positions of the operator in the original constraint text are recorded, and the symbol string of the operator is saved to a temporary variable. The scan continues until the end of the original constraint text, and all the extracted operators are stored in the operator list in the order of their appearance.

[0056] Furthermore, note that when truncating, the original characters at the end of the left substring and the beginning of the right substring are preserved without any trimming. This results in two strings: a left substring and a right substring. The left substring contains the verification field name and possible spaces or other characters, while the right substring contains the verification baseline value and possible spaces or other characters.

[0057] Furthermore, this sequence consists of letters and numbers, which are used as the verification field name for the target process. If no non-alphanumeric characters are encountered during the reverse scan until the beginning of the left substring, the entire left substring is used as the verification field name. However, the left substring may contain spaces. The scan ensures that only alphanumeric strings adjacent to the operator are extracted.

[0058] Furthermore, this sequence consists of numbers and decimal points, which are used as the verification benchmark for the target process. If the first character in the right substring is not a number, a null value may be extracted. However, according to the format of the original constraint text, the first non-space character is usually a number. The scan ensures that the sequence of consecutive numeric characters adjacent to the operator is extracted.

[0059] In summary, this technology can capture not only single-character operators such as greater than, less than, and equal to, but also two-character operators such as greater than or equal to, less than or equal to, and not equal to, ensuring that subsequent logical comparisons accurately reflect the original intent of the rules. Compared to traditional technologies that only support a single comparison method or require manual specification of operators, this traversal extraction mechanism can automatically adapt to expressions of varying complexity, providing accurate operator information for subsequent segmentation and parsing, and avoiding deviations in audit rule execution caused by incomplete or incorrect operator recognition.

[0060] In summary, this method ensures that the two split parts each contain the complete original content of the validation field name and the validation baseline value, while preserving the correspondence between the field name and the baseline value established through operators. Compared to methods using regular expression matching or fixed-position truncation, this segmentation method based on the actual occurrence position can adapt to field names and baseline values ​​of different lengths, as well as spaces that may exist between field names and operators, significantly improving compatibility and parsing success rate for original constraint text of various formats.

[0061] In summary, this method leverages the proximity of field names and operators in the text, starting the search from the position closest to the operator. It automatically skips spaces between field names and operators and filters out irrelevant characters by extracting only alphanumeric strings, ensuring that the extracted validation field names are clean and directly usable for data retrieval. This approach avoids extraction errors caused by special characters or spaces in field names, guaranteeing the accuracy of subsequent searches for actual values ​​in audit event logs.

[0062] In summary, this forward scanning strategy starts from the rightmost position closest to the operator and sequentially collects consecutive numerical characters, including numbers and decimal points. It automatically filters leading spaces from the baseline value and supports threshold extraction for both integer and decimal types. Compared to traditional methods that require manual threshold setting or only support integer comparisons, this scanning mechanism ensures that the verification baseline value retains its original precision, avoiding comparison errors caused by numerical truncation or format conversion, and providing an accurate comparison basis for subsequent risk status label generation.

[0063] In this embodiment of the invention, the step of performing logical operations on the actual value corresponding to the field name to be verified in the audit event record and the verification benchmark value according to the operator, and generating the risk status identifier of the target process based on the operation result, is specifically used for: Extract the actual value that matches the field name to be verified from the audit event record, and compare the actual value with the amount threshold and approval time limit threshold in the verification benchmark value according to the operator to obtain the comparison result value of the target process; If the comparison result is true, the risk level code of the target process is extracted from the process audit constraint, and the current time timestamp of the target process is obtained based on the marker time of the risk level code. The risk level code, the current time timestamp, and the node identifier of the instance audit marker node are combined to form the risk status identifier of the target process. If the comparison result is false, a null value risk status identifier is generated and written into the risk status field of the instance audit marker node.

[0064] Specifically, the audit event log is searched for a field name that is exactly the same as the string of the field name to be verified. The field value corresponding to the field name is extracted as the actual value. The actual value may be an amount or a time value. Then, the amount threshold and the approval time limit threshold are identified from the verification benchmark value. The amount threshold is a numerical value, and the approval time limit threshold is a time length value.

[0065] Specifically, when the comparison result is true, the risk level code field is searched from the process audit constraints. This field stores a predefined risk level identifier, such as high risk, medium risk, or low risk. The value of this field is extracted as the risk level code. Then, the operating system's time service interface is called to obtain the current system date and time. The date and time are converted into a timestamp string according to a fixed format as the current time timestamp.

[0066] Specifically, when the comparison result is false, no risk level code extraction operation is performed. Instead, a null risk status identifier is directly generated. The null risk status identifier is a special string such as an empty string. Then, the audit mark node of the instance is found in the process instance status data, the risk status field of the node is located, and the null risk status identifier is written into the field, overwriting the original content of the field.

[0067] Furthermore, the actual value is compared with the amount threshold using operators, and the actual value is also compared with the approval time limit threshold using operators. If the operator is greater than, it checks whether the actual value is greater than the threshold; if the operator is less than, it checks whether the actual value is less than the threshold; if the operator is equal to, it checks whether the actual value is equal to the threshold; for double-character operators such as greater than or equal to, it checks whether it is greater than or equal to the threshold. The results of the two comparisons are then logically ANDed. That is, the final comparison result is true only when both the amount comparison and the time limit comparison are true; otherwise, it is false. This Boolean comparison result is used as the comparison result value of the target process.

[0068] Furthermore, the node identifier is read from the instance audit marker node, and the three strings of risk level code, current time timestamp and node identifier are concatenated into a complete string according to the preset separator. This string is used as the risk status identifier of the target process. This risk status identifier is used to mark the node as having risk in the audit process view later.

[0069] Furthermore, if the field does not exist, a new risk status field is added to the node data structure and an empty value is written. After the risk status identifier is written, the risk status field of the node is empty, indicating that the node has not triggered a risk.

[0070] In summary, this system achieves automated dual-dimensional verification of key financial data in audit events. By simultaneously comparing two thresholds—amount and approval timeframe—it comprehensively identifies whether operations comply with preset risk control standards, avoiding potential omissions or false alarms that might occur with traditional single-dimensional judgments. The logical comparison process strictly adheres to the definitions of operators, ensuring the accuracy and consistency of comparison results and providing a reliable basis for subsequent risk assessment, thereby improving the precision and coverage of audit verification.

[0071] In summary, key information triggering risks is integrated into a standardized data structure that includes the risk level, occurrence time, and associated node, facilitating intuitive visualization and rapid tracing within the audit process view. By capturing risk marker moments in real time and generating timestamps, the traceability and timeliness of risk status are ensured, providing managers with accurate risk identification data and significantly improving risk response and handling efficiency.

[0072] In summary, ensuring that each instance audit marker node has a clearly defined risk status record, and that the node's risk status field is filled regardless of whether a risk is triggered, maintains data consistency and integrity. The writing of null values ​​avoids subsequent processing anomalies caused by missing fields, providing a unified data foundation for the generation of dynamic audit workflow views. This allows the view to fully present the risk status of all nodes, while also facilitating batch queries and statistical analysis by the system, improving the standardization and usability of audit data.

[0073] In this embodiment of the invention, when arranging the task nodes according to the start and end timestamps of the historical node flow records in the process instance state, and combining the logical connection relationships between task nodes in the historical node flow records, to obtain the flow path of the target process, the specific usage is as follows: Extract the historical node flow record of the target process from the state of the process instance. The historical node flow record includes node identifier, node type, start timestamp and end timestamp. Read the connection arc definition between task nodes from the process definition file corresponding to the process instance state. The connection arc definition includes a source node identifier and a target node identifier. Taking the task node with the earliest start timestamp as the path start node, and according to the binding relationship between the source node identifier and the target node identifier in the connection arc definition, the historical node flow record is traversed to obtain the node sequence of the target process; During the traversal, for task nodes with parallel branches, the child nodes associated with the task node in the process definition file are located based on the parallel gateway identifier of the node type field value in the task node. The child nodes are arranged in the same level according to the order of the start timestamps. For task nodes whose end timestamp is later than the start timestamp of the successor node, the logical relationship defined by the connection arc is used as the connection basis, and the task node is still arranged first and the successor node is arranged second. The node sequence is combined according to time order and logical relationship, and the status information of the processing efficiency of the task node is written into the node data of the corresponding node in the node sequence during the combination to obtain the flow path of the target process.

[0074] Specifically, the database table for the stored process instance status is first accessed, and all historical flow record rows corresponding to the instance are filtered out based on the unique identifier of the target process instance.

[0075] Specifically, based on the process definition identifier recorded in the process instance status, the file system location storing the process definition is located, the file is loaded and its contents are parsed, and the connection arc element that specifically defines the flow relationship between nodes is searched in the process definition file.

[0076] Specifically, first find the record with the smallest start timestamp value from the historical node flow records, take the node identifier of the record as the starting node of the path, and then take the node identifier as the current node. Then, find all connection arcs in the connection arc definition list whose source node identifier is equal to the current node identifier. The target node identifier corresponding to each such connection arc is the direct successor node of the current node.

[0077] Specifically, during the traversal, for task nodes with parallel branches, the system determines whether the node type field value is a parallel gateway identifier. If the node type field value is equal to a predefined parallel gateway identifier string, the system locates the parallel gateway node configuration in the process definition file and reads all child node identifiers defined under that configuration.

[0078] Specifically, the previously obtained node sequence is used as the basic framework, and the nodes are organized into a tree or nested list according to the hierarchical structure of parallel branches, traversing each node during the organization process.

[0079] Furthermore, each record contains four fields: node identifier, node type, start timestamp, and end timestamp. The values ​​of these fields are fully read, sorted in ascending order by the start timestamp, and saved as a list. This list is the historical node flow record of the target process, which fully describes the detailed information of each task node that the process instance has passed through.

[0080] Furthermore, each connection arc element contains two attribute values: a source node identifier and a target node identifier. By traversing all connection arc elements, each pair of source node identifiers and target node identifiers is extracted to form a connection arc definition list. This list reflects the logical flow relationship between task nodes as specified in the process design.

[0081] Next, the records corresponding to these successor node identifiers are found in the historical node flow records, and these records are sorted in ascending order according to their start timestamps. The current node is then stored in the node sequence, and each successor node is used as the new current node. The above search and sorting operations are repeated until no successor node can be found. The final node sequence is arranged in the order of the actual execution time of the process.

[0082] Furthermore, the records corresponding to these child node identifiers are then selected from the historical node flow records. These child node records are sorted in ascending order according to their start timestamps. After sorting, they are inserted as nodes of the same level after the current parallel gateway node. For task nodes encountered during traversal whose end timestamp is later than the start timestamp of the successor node, their order of precedence and successor node is not changed. They are still arranged according to the logical relationship specified in the connection arc definition to ensure that the order of the node sequence is consistent with the process definition.

[0083] Furthermore, the processing efficiency value matching the node identifier is found from the pre-calculated processing efficiency data set, and this processing efficiency value is added as a status information field to the node data object of the node. After all nodes have completed the status information writing, this sequence of nodes with processing efficiency status information constitutes the flow path of the target process. This flow path can intuitively display the process execution order and the efficiency status of each node.

[0084] In summary, by preserving node identifiers, the unique identifiability of each node in the process definition is ensured; by recording node types, a basis for distinguishing between ordinary task nodes and parallel gateway nodes is provided; and by saving start and end timestamps, the actual execution time and relative order of each node can be accurately quantified. This multi-dimensional historical data extraction method overcomes the limitations of traditional technologies that only record node names or execution results, providing complete data support for accurate reconstruction of process paths and efficiency evaluation.

[0085] In summary, by reading the binding relationship between the source node identifier and the target node identifier, the legal flow paths specified in the process design are clarified, such as the direction of sequential branches, conditional branches, or parallel branches. This connection arc definition based on the design document ensures that the subsequently constructed node sequence strictly conforms to the expected logic of the business process, avoiding logical errors that may occur if sorting is based solely on timestamps, such as misjudging nodes that should be executed sequentially as parallel, or incorrectly arranging nodes that should be parallel as sequential.

[0086] In summary, by starting from the earliest node and searching for successor nodes step by step according to the definition of connecting arcs, this method ensures that the generated node sequence not only conforms to the actual execution time order but also strictly follows the logical flow of the process design. This method can automatically handle complex structures such as loops and branches that may exist in the process. Compared with manually sorting logs, it significantly improves the accuracy and efficiency of path reconstruction and provides standardized node sequence input for subsequent process analysis and auditing.

[0087] In summary, by identifying and locating child nodes through the parallel gateway and arranging them in chronological order at the same level, the execution of parallel branches is accurately reproduced, avoiding the erroneous treatment of parallel nodes as sequential nodes. For cases where the termination timestamp is later than the start timestamp of the successor node, the logical relationship defined by the connection arc is adhered to for arrangement, avoiding incorrect adjustment of the node order due to time overlap. This ensures that even when there is time overlap in actual execution, the node sequence can still accurately reflect the logical flow of the process design.

[0088] In summary, by embedding processing efficiency status information into the node data of each node, the final workflow path not only displays the execution order and parallel structure of the nodes but also intuitively reflects the efficiency performance of each node, such as high efficiency, normal, or low efficiency. This workflow path, which integrates efficiency assessment, provides managers with a decision-making basis for analyzing bottleneck nodes and optimizing resource allocation from a global process perspective. It overcomes the limitation of traditional process views that only display node order and lack performance data, significantly improving the depth and practicality of audit process management.

[0089] In this embodiment of the invention, the formula for calculating the processing efficiency is specifically used for: in, For the first The processing efficiency coefficient of each node. For node indexing, Based on basic efficiency weights, For the first Standard processing time configuration parameters for each node For the first The actual time each node is occupied To deviate from the penalty weight, It is a natural constant. To await the penalty weight, It is the natural logarithm function. For the first The idle time of each node.

[0090] Specifically, the standard processing time configuration parameter is derived from the expected processing time predefined for each task node in the process definition file. This time is a fixed duration value set during the process design phase based on business experience or historical statistical data. It is read from the process definition file and stored in the node configuration information when the process instance starts. The actual occupied time is derived from the time difference between the end timestamp and the start timestamp of the node recorded in the historical node flow record. This time difference accurately reflects the actual time consumed by the node from the start of execution to completion. The flow idle time is derived from the time difference between the start timestamp and the end timestamp of the previous adjacent node in the historical node flow record. If the node is the starting node of the path, the flow idle time is zero. This time difference represents the time spent by the node waiting for preconditions to be met before entering the execution state. The three weight values—basic efficiency weight, deviation penalty weight, and waiting penalty weight—are derived from the system configuration parameter file and are fixed constants loaded during the initialization of the audit process processing system. These weight values ​​remain unchanged throughout the entire audit process processing and are used to adjust the influence of the three components in the calculation of the processing efficiency coefficient. Natural constants are fixed constants in mathematics, and their predefined values ​​are used directly in calculations.

[0091] Furthermore, the significance of the formula lies in comprehensively evaluating the processing efficiency of a task node through the combined operation of three parts. The first part divides the standard processing time by the actual occupied time; the ratio reflects the node's basic efficiency in time consumption. A larger ratio indicates a faster processing speed. The second part squares the deviation between the actual occupied time and the standard processing time, takes the negative value, and then performs an exponential operation with the natural constant as the base. The result is a value between zero and one. When the actual occupied time equals the standard processing time, this part reaches its maximum value of one; when the actual occupied time deviates from the standard processing time, this part decreases, with the value decreasing as the deviation increases. This part reflects the penalty for deviating from the standard time. The third part first divides the idle time by the standard processing time to obtain the relative idle time. Then, it adds one to this relative idle time, takes the natural logarithm, and takes the reciprocal of the result. The final result is a value that decreases as the idle time increases; this part reflects the penalty for idle waiting time on processing efficiency. The three parts are multiplied by their respective weights and then added together. The final processing efficiency coefficient comprehensively reflects the node's overall performance in three dimensions: time consumption, standard deviation, and idle waiting.

[0092] In summary, the formula's trend manifests as a specific pattern in the processing efficiency coefficient as the three independent variables change. When the actual processing time is less than the standard processing time, the ratio of the standard time to the actual time in the first part is greater than one, resulting in a larger value for the first part. Simultaneously, in the second part, the difference between the actual time and the standard time is negative; squaring this result yields a positive number, but the exponential result is less than one but greater than zero. The combined effect of these two parts leads to a higher overall efficiency coefficient. When the actual processing time exceeds the standard processing time, the ratio of the standard time to the actual time in the first part is less than one, resulting in a smaller value for the first part. Simultaneously, in the second part, the square of the positive result of the actual time minus the standard time, followed by a negative value, significantly reduces the exponential result. Both parts contribute to a decrease in the overall efficiency coefficient, with the decrease becoming more pronounced as the actual time exceeds the standard time. When the actual processing time equals the standard processing time, the second part reaches its maximum value, and the efficiency coefficient reaches its highest value for that actual processing time. The impact of idle time on the efficiency coefficient shows a monotonically decreasing trend. That is, the longer the idle time, the smaller the value of the third part. Due to the characteristics of the natural logarithm function, the efficiency coefficient decreases rapidly in the early stage of increasing idle time. As the idle time continues to increase, the rate of decrease of the efficiency coefficient gradually slows down and eventually approaches zero but will never be equal to zero.

[0093] In this embodiment of the invention, when attaching the risk status identifier to the corresponding node of the flow path to obtain a dynamic audit process view of the target process, it is specifically used for: Based on the value of the risk status identifier, the corresponding nodes of the flow path are marked to obtain the initial audit process view of the target process; The trigger timestamp of the risk status identifier is converted into a standard date and time format and written into the remarks field of the corresponding node in the initial audit process view to obtain the dynamic audit process view of the target process.

[0094] Specifically, first, each node in the flow path is traversed, and the previously stored risk status identifier field is read from the data object of that node. The value status of the field is checked. If the value of the risk status identifier field is an empty risk status identifier, no marking processing is performed on that node. If the value of the risk status identifier field is a non-empty risk status identifier string, the risk level code part is parsed from the string, and the marking type is determined according to the value of the risk level code.

[0095] Specifically, the current time stamp portion is extracted from each non-empty risk status identifier string. The current time stamp is a numeric string in seconds or milliseconds. This string is passed to the date and time formatting function. The function rearranges the numbers according to a predefined standard format, such as four-digit year, two-digit month, two-digit day, two-digit hour, two-digit minute, and two-digit second, and adds hyphens and colons as separators between each part to obtain a string in standard date and time format.

[0096] Furthermore, in the graphical representation of the flow path, the corresponding graphical element of the node is found, and the graphical rendering interface is called to modify the border color or fill color of the graphical element to the corresponding mark color. At the same time, a risk icon is added next to the graphical element. After the marking is completed, each node in the flow path is correctly colored and marked according to the value of its risk status identifier. This flow path with color and icon is the initial audit process view of the target process.

[0097] Furthermore, in the initial audit process view, locate the node graphic element corresponding to the risk status identifier, find the storage location of the note field of the node. The note field usually exists in the form of a text box or a hidden attribute. Write the converted standard date and time string into the note field. If the note field originally has content, append it after the original content. After writing, the note field of the node will contain the risk trigger time information. After the trigger time of all risk nodes has been written, the initial audit process view is updated to the dynamic audit process view. This view not only displays the risk mark of the node, but also provides the specific trigger time of each risk node.

[0098] In summary, by assigning labels based on risk level codes—for example, high risk in red, medium risk in yellow, and low risk in blue—managers can identify high-risk stages in the process immediately without having to review detailed data for each node. This visual risk alert significantly improves the intuitiveness and readability of audit results, solving the problem of traditional audit reports that only list risks in text form, making it difficult to quickly locate them. At the same time, leaving non-risk nodes unmarked maintains a clear and concise view, avoiding interference from redundant information. The generated initial audit process view, as the first fusion of risk status and process path, provides a structured view framework for subsequently adding time information.

[0099] In summary, by converting the original timestamp strings into standard date and time formats such as four-digit year, two-digit month, two-digit day, hour, minute, and second, the readability and consistency of time information are ensured, making it easier for auditors to understand the specific moment when a risk occurred. The converted time is written to the remarks field of the corresponding node, so that each risk node not only carries a risk level marker but also includes traceable time evidence. When reviewing a risk event at a particular node, the accurate occurrence time can be directly obtained from the remarks field without needing to trace back to the original logs. The resulting dynamic audit process view fully integrates information from four dimensions: node sequence, logical relationships, risk level, and trigger time, providing comprehensive and accurate visual data support for audit traceability and process optimization in financial management.

[0100] Compared with the prior art, the present invention has the following beneficial effects:

[0101] 1. This technical solution achieves high-precision correlation between audit events and process nodes by capturing data change records in real time from the database transaction log and comparing them with the dynamic running status of process instances through time-series comparison and feature anchoring. This dynamic correlation mechanism ensures that every business operation can be accurately assigned to a specific process stage, so that audit analysis no longer relies on manual retrospective tracing, but can obtain complete operation records of each node during process execution. Simultaneously, the solution automatically extracts verification field names, operators, and verification benchmark values ​​by dynamically parsing the node-level constraint text in the process definition file, achieving deep binding between audit rules and process design. When the business process is adjusted, only the constraints in the definition file need to be modified to update the audit logic synchronously, significantly improving the flexibility and responsiveness of rule configuration and ensuring that audit verification is always consistent with the latest business requirements.

[0102] 2. This technical solution generates a dynamic audit process view rich in multi-source information by integrating time, logic, and efficiency dimensions. In reconstructing the process flow path, the solution not only accurately restores the execution order and parallel structure of nodes based on timestamps and connection arcs, but also embeds processing efficiency status information into the data of each node, enabling the process view to intuitively display the efficiency performance of each node. The value status of risk status identifiers is further transformed into visual node markers, and the trigger time is converted into a standard format and written into the remarks field. The final dynamic audit process view fully presents the node sequence, logical relationships, risk levels, and processing efficiency. This multi-dimensional visualization output provides managers with comprehensive process insights, enabling precise data support for bottleneck node identification, risk positioning, and resource allocation optimization, significantly improving the refinement level of audit process management.

[0103] like Figure 2 The diagram shown is a functional block diagram of an audit process processing system for financial management provided by an embodiment of the present invention.

[0104] The audit process processing system 100 for financial management described in this invention can be installed in an electronic device. Depending on the functions implemented, the audit process processing system 100 for financial management may include a data configuration module 101, an instance audit tagging module 102, a process audit constraint module 103, a risk status identification module 104, a workflow path module 105, and a dynamic audit process view module 106. The modules described in this invention can also be referred to as units, which are a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, stored in the memory of the electronic device.

[0105] In this embodiment, the functions of each module / unit are as follows: The data configuration module extracts the audit event records and current process status data of the target process based on the business operation events and process instance status of the target process. The instance audit tagging module associates and maps the audit event record with the current dynamic node of the current state data of the process to obtain the instance audit tagging node of the target process; The process audit constraint module locates the process audit constraint corresponding to the instance audit marker node based on the process definition file corresponding to the process instance status. The process audit constraint includes the field name to be verified, the operator, and the verification benchmark value. The risk status identification module performs logical operations on the actual value corresponding to the field name to be verified in the audit event record and the verification benchmark value according to the operator, and generates the risk status identification of the target process based on the operation result. The flow path module arranges the task nodes according to the start and end timestamps of the historical node flow records in the process instance status, and in combination with the logical connection relationship between the task nodes in the historical node flow records, to obtain the flow path of the target process. The dynamic audit process view module attaches the risk status identifier to the corresponding node of the flow path to obtain a dynamic audit process view of the target process.

[0106] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0107] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0108] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0109] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0110] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0111] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for handling audit processes applicable to financial management, characterized in that, The method includes: Based on the business operation events and process instance status of the target process, extract the audit event records and current process status data of the target process; The audit event record is associated with the current dynamic node of the current state data of the process to obtain the instance audit marker node of the target process; Based on the process definition file corresponding to the process instance status, the process audit constraint corresponding to the instance audit marker node is located. The process audit constraint includes the field name to be verified, the operator, and the verification benchmark value. The actual value corresponding to the field name to be verified in the audit event record is logically operated on with the verification benchmark value according to the operator, and a risk status identifier of the target process is generated based on the operation result; Based on the start and end timestamps of the historical node flow records in the process instance status, and combined with the logical connection relationships between task nodes in the historical node flow records, the task nodes are arranged to obtain the flow path of the target process. The risk status identifier is attached to the corresponding node of the flow path to obtain a dynamic audit process view of the target process.

2. The audit process handling method applicable to financial management as described in claim 1, characterized in that, The step of extracting audit event records and current process status data of the target process based on business operation events and process instance status includes: Data change records are captured from the database transaction log of the target process, and the operation type identifier, operation timestamp, operator account name and financial data field value in the data change records are combined into the business operation event of the target process; Traverse the process instance states in the target process, extract the current node identifier, node trigger timestamp, associated business data index, and node traversal trajectory of the business instance in the process instance state of the target process, and obtain the current process state data of the target process; The operation timestamp of the business operation event is compared with the node trigger timestamp. If the operation timestamp is later than the node trigger timestamp, the business operation event is recorded as an audit event of the target process. If the operation timestamp is earlier than or equal to the node trigger timestamp, the node with a termination timestamp later than the operation timestamp is searched from the node's experience trajectory, and the business operation event is assigned to the audit event record of the node.

3. The audit process handling method applicable to financial management as described in claim 1, characterized in that, The step of associating and mapping the audit event record with the current dynamic node of the current state data of the process to obtain the instance audit marker node of the target process includes: Align the timestamp of the audit event record with the time point of the status change of each dynamic node in the current status data of the process, and filter out the target dynamic node that matches the time of the audit event record. The event type identifier of the audit event record is compared with the node attribute information of the target dynamic node. Audit event records with matching characteristics are anchored to the target dynamic node to generate an instance audit marker node for the target process.

4. The audit process handling method applicable to financial management as described in claim 1, characterized in that, The process audit constraints corresponding to the instance audit marker node are located based on the process definition file corresponding to the process instance state. These process audit constraints include the field names to be verified, operators, and verification benchmark values, including: Based on the process instance status, the process definition identifier reads the corresponding process definition file, and matches the corresponding node configuration information segment in the process definition file according to the node identifier of the instance audit mark node. Scan the condition expression field in the node configuration information segment to obtain the original constraint text of the target process; The original constraint text is parsed to identify the verification field names, operators, and verification benchmark values ​​in the original constraint text, thereby obtaining the process audit constraints of the target process.

5. The audit process handling method applicable to financial management as described in claim 4, characterized in that, The process of parsing the original constraint text to identify the validation field names, operators, and validation benchmark values ​​in the original constraint text, and obtaining the process audit constraints of the target process, includes: Traverse the character sequence of the original constraint text and extract the operators in the character sequence; Using the position of the first relational operation character in the original constraint text as the dividing point, the original constraint text is divided into a left substring and a right substring; Starting from the end of the left substring, scan backwards and use the alphanumeric string adjacent to the relational operation character as the verification field name of the target process; Starting from the first character of the right substring, scan sequentially backwards, and use the consecutive numerical character sequence adjacent to the relational operation character as the verification benchmark value of the target process.

6. The audit process handling method applicable to financial management as described in claim 1, characterized in that, The step of performing logical operations on the actual value corresponding to the field name to be verified in the audit event record and the verification benchmark value according to the operator, and generating a risk status identifier for the target process based on the operation result, includes: Extract the actual value that matches the field name to be verified from the audit event record, and compare the actual value with the amount threshold and approval time limit threshold in the verification benchmark value according to the operator to obtain the comparison result value of the target process; If the comparison result is true, the risk level code of the target process is extracted from the process audit constraint, and the current time timestamp of the target process is obtained based on the marker time of the risk level code. The risk level code, the current time timestamp, and the node identifier of the instance audit marker node are combined to form the risk status identifier of the target process. If the comparison result is false, a null value risk status identifier is generated and written into the risk status field of the instance audit marker node.

7. The audit process handling method applicable to financial management as described in claim 1, characterized in that, The step of arranging the task nodes according to the start and end timestamps of the historical node flow records in the process instance status, and combining the logical connection relationships between task nodes in the historical node flow records, to obtain the flow path of the target process includes: Extract the historical node flow record of the target process from the state of the process instance. The historical node flow record includes node identifier, node type, start timestamp and end timestamp. Read the connection arc definition between task nodes from the process definition file corresponding to the process instance state. The connection arc definition includes a source node identifier and a target node identifier. Taking the task node with the earliest start timestamp as the path start node, and according to the binding relationship between the source node identifier and the target node identifier in the connection arc definition, the historical node flow record is traversed to obtain the node sequence of the target process; During the traversal, for task nodes with parallel branches, the child nodes associated with the task node in the process definition file are located based on the parallel gateway identifier of the node type field value in the task node. The child nodes are arranged in the same level according to the order of the start timestamps. For task nodes whose end timestamp is later than the start timestamp of the successor node, the logical relationship defined by the connection arc is used as the connection basis, and the task node is still arranged first and the successor node is arranged second. The node sequence is combined according to time order and logical relationship, and the status information of the processing efficiency of the task node is written into the node data of the corresponding node in the node sequence during the combination to obtain the flow path of the target process.

8. The audit process processing method applicable to financial management as described in claim 7, characterized in that, The formula for calculating the processing efficiency includes: in, For the first The processing efficiency coefficient of each node. For node indexing, Based on basic efficiency weights, For the first Standard processing time configuration parameters for each node For the first The actual time each node is occupied To deviate from the penalty weight, It is a natural constant. To await the penalty weight, It is the natural logarithm function. For the first The idle time of each node.

9. The audit process handling method applicable to financial management as described in claim 1, characterized in that, The step of attaching the risk status identifier to the corresponding node of the flow path to obtain a dynamic audit process view of the target process includes: Based on the value of the risk status identifier, the corresponding nodes of the flow path are marked to obtain the initial audit process view of the target process; The trigger timestamp of the risk status identifier is converted into a standard date and time format and written into the remarks field of the corresponding node in the initial audit process view to obtain the dynamic audit process view of the target process.

10. An audit process processing system suitable for financial management, used to implement the audit process processing method suitable for financial management as described in any one of claims 1-9, characterized in that, The system includes: The data configuration module extracts the audit event records and current process status data of the target process based on the business operation events and process instance status of the target process. The instance audit tagging module associates and maps the audit event records with the current dynamic nodes of the current state data of the process to obtain the instance audit tagging node of the target process; The process audit constraint module locates the process audit constraint corresponding to the instance audit marker node based on the process definition file corresponding to the process instance status. The process audit constraint includes the field name to be verified, the operator, and the verification benchmark value. The risk status identification module performs logical operations on the actual value corresponding to the field name to be verified in the audit event record and the verification benchmark value according to the operator, and generates the risk status identification of the target process based on the operation result. The flow path module arranges the task nodes according to the start and end timestamps of the historical node flow records in the process instance status, and in combination with the logical connection relationship between the task nodes in the historical node flow records, to obtain the flow path of the target process. The dynamic audit process view module attaches the risk status identifier to the corresponding node of the flow path to obtain a dynamic audit process view of the target process.

Images