Process compliance examination method and device, equipment and storage medium

Abstract

The invention discloses a process compliance review method, device and equipment and a storage medium, and belongs to the technical field of process compliance review. The method comprises the following steps: firstly, performing data preprocessing on to-be-reviewed text data to obtain processed to-be-reviewed text data; constructing an examination logic database based on a pre-trained AI model and the processed to-be-examined text data; and performing compliance review detection on the processed to-be-reviewed text data based on the review logic database to obtain a corresponding process compliance review result. According to the invention, compliance review detection verification is carried out on the processed to-be-reviewed text data through the pre-trained AI model, intelligentization and automation of business system auditing are realized, and the working efficiency and accuracy of process compliance review are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of process compliance review, and in particular to a process compliance review method, device, equipment and storage medium. Background Art

[0002] The review of project compliance covers the entire life cycle of a project from initiation to completion. By conducting in-depth inspections of the standardization, consistency and completeness of the documents output at each stage, the compliance management of the project process can be effectively achieved.

[0003] In the existing technology, whether the logic between project documents is often required to be manually checked, such as whether contracts, reimbursement forms, payment forms, invoices, acceptance forms, delivery notes, settlement forms, etc. meet the requirements of the preset system, and whether the relationship between them is logical. This requires organizing manpower to extract valuable information and associations from huge project documents one by one. When the project is complex and the number of documents is too large, it often takes several months to check. Even if computer technology regular expression technology is introduced, it usually relies on static rules, such as using regular expressions to match keywords, which makes it difficult to accurately extract multiple types of (such as cause and effect, condition, time sequence, parallel, etc.) event logic relationships from unstructured texts such as policy documents and business processes. It is also impossible to efficiently detect complex logical violations, such as missing preconditions, conflicting cause and effect relationships, vague descriptions of fund use, illegal misappropriation of funds, unreasonable expense expenditure instructions, and so on.

[0004] In view of this, there is an urgent need to obtain a process compliance review method, device, equipment and storage medium to improve the work efficiency and accuracy of process compliance review.

[0005] The above contents are only used to assist in understanding the technical solution of the present invention and do not constitute an admission that the above contents are prior art. Summary of the invention

[0006] The main purpose of the present invention is to provide a process compliance review method, device, equipment and storage medium, aiming to solve the technical problem of how to improve the work efficiency and accuracy of process compliance review in the prior art.

[0007] To achieve the above object, the present invention provides a process compliance review method, which comprises the following steps: Performing data preprocessing on the text data to be reviewed to obtain processed text data to be reviewed; Building a review logic database based on the pre-trained AI model and the processed text data to be reviewed; Based on the review logic database, the processed text data to be reviewed is subjected to compliance review detection to obtain corresponding process compliance review results.

[0008] Optionally, constructing a review logic database based on the pre-trained AI model and the processed text data to be reviewed specifically includes: Performing data analysis on the processed text data to be reviewed based on the pre-trained AI model to obtain event structured data; Generate a vector for the processed text data to be reviewed based on the pre-trained AI model to obtain an event semantic vector; A review logic database is constructed based on the event structured data and the event semantic vector.

[0009] Optionally, constructing a review logic database based on the event structured data and the event semantic vector specifically includes: Intelligently analyzing the event structured data and the event semantic vector through the pre-trained AI model to obtain an intelligent analysis result; Performing data processing on the intelligent analysis results to obtain a processed intelligent analysis database; The output format of the processed intelligent analysis database is controlled through a standardized prompting project to obtain a review logic database.

[0010] Optionally, data processing is performed on the intelligent analysis result to obtain a processed intelligent analysis database, which specifically includes: Perform event relationship verification based on the intelligent analysis result to obtain a verified intelligent analysis data set; Completing the relationship description based on the intelligent parsing result to obtain a completed intelligent parsing data set; The verified intelligent analysis data set and the completed intelligent analysis data set are integrated to obtain a processed intelligent analysis database.

[0011] Optionally, a compliance review is performed on the processed text data to be reviewed based on the review logic database to obtain a corresponding process compliance review result, specifically including: Based on the review logic database, the processed text data to be reviewed is mapped and compared to obtain a data mapping comparison result; Based on dynamic rule data, the data mapping comparison results are checked for violations to obtain corresponding process compliance review results.

[0012] Optionally, the data mapping comparison result is detected for violations based on the dynamic rule data to obtain corresponding process compliance review results, including: Perform violation detection on data mapping comparison results based on dynamic rule data; If the violation detection is qualified, the normal process compliance review result is obtained; If the violation detection is unqualified, the abnormal process compliance review result is obtained and the corresponding violation review report is output.

[0013] Optionally, the dynamic rule data generation method may include manual setting and entry or automatic generation by the pre-trained AI model based on process review rules.

[0014] In addition, to achieve the above-mentioned purpose, the present invention also proposes a process compliance review device, the process compliance review device comprising: Data processing module: performs data preprocessing on the text data to be reviewed to obtain the processed text data to be reviewed; Database construction module: constructing a review logic database based on the pre-trained AI model and the processed text data to be reviewed; Process review module: Based on the review logic database, the processed text data to be reviewed is subjected to compliance review and detection to obtain corresponding process compliance review results.

[0015] In addition, to achieve the above-mentioned purpose, the present invention also proposes a process compliance review device, which includes: a memory, a processor, and a process compliance review program stored on the memory and executable on the processor, and the process compliance review program is configured to implement the steps of the process compliance review method described above.

[0016] In addition, to achieve the above-mentioned purpose, the present invention also proposes a computer-readable storage medium storing a computer program, wherein a process compliance review program is stored on the storage medium, and when the process compliance review program is executed by a processor, the steps of the process compliance review method described above are implemented.

[0017] The present invention first performs data preprocessing on the text data to be reviewed to obtain the processed text data to be reviewed; constructs a review logic database based on the pre-trained AI model and the processed text data to be reviewed; performs compliance review detection on the processed text data to be reviewed based on the review logic database to obtain the corresponding process compliance review results. The present invention performs compliance review detection and verification on the processed text data to be reviewed through a pre-trained AI model, realizes the intelligentization and automation of business system auditing, and improves the work efficiency and accuracy of process compliance review. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a structural diagram of a process compliance review device of a hardware operating environment involved in an embodiment of the present invention; Figure 2 It is a flowchart of the first embodiment of the process compliance review method of the present invention; Figure 3 It is a flowchart of the second embodiment of the process compliance review method of the present invention; Figure 4 It is a flowchart of the third embodiment of the process compliance review method of the present invention; Figure 5 It is a structural block diagram of the first embodiment of the process compliance review device of the present invention.

[0019] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0020] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0021] Reference Figure 1 , Figure 1 A schematic diagram of the structure of a process compliance review device for a hardware operating environment according to an embodiment of the present invention.

[0022] like Figure 1 As shown, the process compliance review device may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the user interface 1003 may also include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a wireless fidelity (Wireless-Fidelity, Wi-Fi) interface). The memory 1005 may be a high-speed random access memory (Random Access Memory, RAM) or a stable non-volatile memory (Non-Volatile Memory, NVM), such as a disk storage. The memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0023] Those skilled in the art will understand that Figure 1 The structure shown in the figure does not constitute a limitation on the process compliance review device, and may include more or less components than shown in the figure, or combine certain components, or arrange the components differently.

[0024] like Figure 1As shown, the memory 1005 as a storage medium may include an operating system, a network communication module, a user interface module, and a process compliance review program.

[0025] exist Figure 1 In the process compliance review device shown, the network interface 1004 is mainly used for data communication with the network server; the user interface 1003 is mainly used for data interaction with the user; the processor 1001 and the memory 1005 in the process compliance review device of the present invention can be set in the process compliance review device, and the process compliance review device calls the process compliance review program stored in the memory 1005 through the processor 1001, and executes the process compliance review method provided by the embodiment of the present invention.

[0026] The present invention provides a process compliance review method, referring to Figure 2 , Figure 2 The figure is a flowchart of a first embodiment of a process compliance review method of the present invention.

[0027] In this embodiment, the process compliance review method includes the following steps: Step S10: preprocessing the text data to be reviewed to obtain processed text data to be reviewed; It should be noted that, in the specific implementation, the role of data preprocessing of the text data to be reviewed is to eliminate errors, inconsistencies and redundancies in the original data through operations such as data cleaning, denoising, and missing value processing; or to integrate data from different sources or forms into a unified format through operations such as data integration, conversion, standardization, and normalization, so as to facilitate subsequent pre-trained AI model processing.

[0028] It should also be noted that, in the specific implementation, the text data to be reviewed includes unstructured data such as policy documents, business process texts, log files, etc., and its data format categories include text format, PDF format, and text format after image text recognition, etc.

[0029] Step S20: Building a review logic database based on the pre-trained AI model and the processed text data to be reviewed; It should be noted that in the specific implementation, the pre-trained AI model is trained based on the Deepseek_R1 large model, which can be used to extract corresponding event and logical relationship data from the text data to be reviewed, and then build a review logic database.

[0030] It should also be noted that in a specific implementation, the review logic database may include a transaction graph and a vector retrieval library, wherein the transaction graph is composed of a Neo4j graph database and the vector retrieval library is composed of a FAISS vector library.

[0031] Step S30: Perform compliance review on the processed text data to be reviewed based on the review logic database to obtain corresponding process compliance review results.

[0032] It should be noted that, in the specific implementation, the process compliance review results can be used to reflect the compliance of this process and output the corresponding compliance detection report.

[0033] This embodiment first performs data preprocessing on the text data to be reviewed to obtain the processed text data to be reviewed; constructs a review logic database based on the pre-trained AI model and the processed text data to be reviewed; performs compliance review detection on the processed text data to be reviewed based on the review logic database to obtain the corresponding process compliance review results. This embodiment performs compliance review detection and verification on the processed text data to be reviewed by the pre-trained AI model, realizes the intelligentization and automation of business system auditing, and improves the work efficiency and accuracy of process compliance review.

[0034] refer to Figure 3 , Figure 3 This is a flow chart of a second embodiment of a process compliance review method of the present invention.

[0035] Based on the above first embodiment, in this embodiment, step S20 specifically includes: Step S21: parsing the processed text data to be reviewed based on the pre-trained AI model to obtain event structured data; It should be noted that, in the specific implementation, the specific process of data analysis of the processed text data to be reviewed based on the pre-trained AI model refers to the analysis of the input text data based on the chain of thought technology CoT (Chain of Thought) of the Deepseek_R1 large model, and then the extraction of event descriptions and standardized logical relationships, and finally the output of structured data. The chain of thought technology CoT (Chain of Thought) of the Deepseek_R1 large model can also be used to generate reasoning paths for event logical relationships and complete missing logical relationships.

[0036] It should also be noted that, in a specific implementation, the standardized logical relationship types may include sequential relationship (SEQUENCE) or causal relationship (CAUSES), transition relationship (CONTRAST), conditional relationship (CONDITION), concurrent relationship (CONCURRENT), composition relationship (COMPOSITION), etc.

[0037] Step S22: Generate a vector for the processed text data to be reviewed based on the pre-trained AI model to obtain an event semantic vector; It can be understood that, in the specific implementation, the specific process of generating vectors for the processed text data to be reviewed based on the pre-trained AI model refers to generating vectors for the input text data based on the FAISS of the Deepseek_R1 large model combined with the Embeddings technology (vector embedding technology) to obtain the corresponding event semantic vector.

[0038] Step S23: Construct a review logic database based on event structured data and event semantic vectors.

[0039] It should also be noted that, in the specific implementation, the specific process of building an audit logic database based on event structured data and event semantic vectors is to first write the event structured data into the Neo4j (a high-performance graph engine) graph database to build a matter graph; then use the Embeddings technology to generate event semantic vectors and store them in the FAISS vector library; finally, integrate the matter graph and the FAISS vector library to obtain the audit logic database.

[0040] This embodiment first uses a pre-trained AI model to perform data analysis and vector generation on the processed text data to be reviewed, and obtains event structured data and event semantic vectors accordingly, and then builds a review logic database based on the event structured data and event semantic vectors, providing a basis for subsequent process compliance review.

[0041] Furthermore, a review logic database is constructed based on the event structured data and the event semantic vector, specifically including: performing intelligent analysis on the event structured data and the event semantic vector through a pre-trained AI model to obtain an intelligent analysis result; performing data processing on the intelligent analysis result to obtain a processed intelligent analysis database; and controlling the output format of the processed intelligent analysis database through a standardized prompt project to obtain a review logic database.

[0042] It should be noted that in the specific implementation, in order to ensure the accuracy and completeness of subsequent data, it is necessary to first control the output format of the processed intelligent parsing database through a standardized prompt project (for example, standardized JSON).

[0043] Furthermore, data processing is performed on the intelligent analysis results to obtain a processed intelligent analysis database, which specifically includes: verifying event relationships based on the intelligent analysis results to obtain a verified intelligent analysis data set; completing relationship descriptions based on the intelligent analysis results to obtain a completed intelligent analysis data set; integrating the verified intelligent analysis data set and the completed intelligent analysis data set to obtain a processed intelligent analysis database.

[0044] It should be noted that, in the specific implementation, event relationship verification based on the intelligent parsing results actually refers to the verification of the extracted JSON format data; and relationship description completion based on the intelligent parsing results actually refers to the completion of the relationship description of the extracted intelligent parsing results, that is, through event relationship verification and completion, it is ensured that each event contains the necessary description fields and legal relationship types.

[0045] refer to Figure 4 , Figure 4 This is a flowchart of a third embodiment of a process compliance review method of the present invention.

[0046] Based on the above first embodiment, in this embodiment, step S30 specifically includes: Step S31: mapping and comparing the processed text data to be reviewed based on the review logic database to obtain data mapping and comparison results; It should be noted that, in the specific implementation, the process of mapping and comparing the processed text data to be reviewed based on the review logic database specifically refers to mapping the text data to be reviewed (for example, the business process or operation steps submitted by the user) to the event nodes in the review logic database (matter map), and then obtaining the data mapping and comparison results.

[0047] Step S32: Perform violation detection on the data mapping comparison result based on the dynamic rule data to obtain the corresponding process compliance review result.

[0048] It should be noted that, in the specific implementation, violation detection of data mapping comparison results based on dynamic rule data is essentially to use the Cypher query language in the pre-trained AI model to query the logical relationship between adjacent events in the data mapping comparison results in real time (to detect whether there are necessary preconditions, causal relationships or other business rule requirements), and then obtain the corresponding process compliance review results.

[0049] It should also be noted that in the specific implementation, dynamic rule data is essentially compliance process rules or requirement data, which can be generated through manual input or automatically generated by a preset AI model (such as the Deepseek_R1 large model) according to basic audit rules. For example: in the audit field, the dynamic rule can be set to, if there is a payment approval event, there must be a "pre-conditional goods acceptance event" and the acceptance status must be "passed".

[0050] This embodiment first maps and compares the processed text data to be reviewed based on the review logic database to obtain data mapping comparison results, and then performs violation detection on the data mapping comparison results based on dynamic rule data to obtain corresponding process compliance review results, thereby realizing the qualification review of process data.

[0051] Furthermore, based on the dynamic rule data, the data mapping comparison results are checked for violations to obtain corresponding process compliance review results, specifically including: based on the dynamic rule data, the data mapping comparison results are checked for violations; if the violation detection is qualified, the normal process compliance review results are obtained; if the violation detection is unqualified, the abnormal process compliance review results are obtained, and the corresponding violation review report is output.

[0052] It is understandable that, in specific implementations, normal process compliance review results can only be obtained when the process data for this audit (such as policy documents, business process texts, log files, etc.) fully complies with the dynamic rule data (process compliance rules or requirements).

[0053] It should also be noted that in the specific implementation, the content of the violation review report may include the specific violation type of the audit process data, the corresponding event chain, risk level and handling suggestions, etc., in order to support manual review and subsequent data tracing.

[0054] For example, # Audit report on misappropriation of project funds ## Violation 001 - **Type**: Key process missing; - **Event chain**: Sign the purchase contract → Submit payment approval → Execute transfer; - **Missing link**: cargo acceptance; - **Association rules**: R1-001 (payment requires prior acceptance); - **Risk Level**: High risk; - **Handling suggestions**: 1. Check the goods receipt record of Supplier A; 2. Review the operation log of payment approver Zhang; 3. Freeze the subsequent payment process that has not been accepted; ## Associated risk items - Found similar historical cases: 2023-09 Supplier B false contract case (similarity 92%); Furthermore, the dynamic rule data generation method may include manual setting and input or automatic generation of pre-trained AI models based on process review rules.

[0055] In addition, an embodiment of the present invention further proposes a computer-readable storage medium storing a computer program, wherein a process compliance review program is stored on the storage medium, and when the process compliance review program is executed by a processor, the steps of the process compliance review method described above are implemented.

[0056] Since the storage medium adopts all the technical solutions of all the above embodiments, it has at least all the beneficial effects brought by the technical solutions of the above embodiments, which will not be described one by one here.

[0057] Reference Figure 5 , Figure 5 It is a structural block diagram of the first embodiment of the process compliance review device of the present invention.

[0058] like Figure 5 As shown, the process compliance review device proposed in the embodiment of the present invention includes: Data processing module 10: performs data preprocessing on the text data to be reviewed to obtain processed text data to be reviewed; Database construction module 20: constructing a review logic database based on the pre-trained AI model and the processed text data to be reviewed; Process review module 30: performs compliance review on the processed text data to be reviewed based on the review logic database to obtain corresponding process compliance review results.

[0059] This embodiment first performs data preprocessing on the text data to be reviewed to obtain the processed text data to be reviewed; constructs a review logic database based on the pre-trained AI model and the processed text data to be reviewed; performs compliance review detection on the processed text data to be reviewed based on the review logic database to obtain the corresponding process compliance review results. This embodiment performs compliance review detection and verification on the processed text data to be reviewed by the pre-trained AI model, realizes the intelligentization and automation of business system auditing, and improves the work efficiency and accuracy of process compliance review.

[0060] It should be understood that the above is only an example and does not constitute any limitation on the technical solution of the present invention. In specific applications, technicians in this field can make settings as needed, and the present invention does not limit this.

[0061] It should be noted that the workflow described above is merely illustrative and does not limit the scope of protection of the present invention. In practical applications, technicians in this field can select part or all of them according to actual needs to achieve the purpose of the present embodiment, and no limitation is made here.

[0062] In addition, for technical details that are not described in detail in this embodiment, please refer to the process compliance review method provided in any embodiment of the present invention, and will not be repeated here.

[0063] In addition, it should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.

[0064] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0065] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as a read-only memory (ROM) / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0066] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A process compliance review method, characterized in that: include: Performing data preprocessing on the text data to be reviewed to obtain processed text data to be reviewed; Building a review logic database based on the pre-trained AI model and the processed text data to be reviewed; Based on the review logic database, the processed text data to be reviewed is subjected to compliance review detection to obtain corresponding process compliance review results.

2. The process compliance review method according to claim 1, characterized in that: Building a review logic database based on the pre-trained AI model and the processed text data to be reviewed includes: Performing data analysis on the processed text data to be reviewed based on the pre-trained AI model to obtain event structured data; Generate a vector for the processed text data to be reviewed based on the pre-trained AI model to obtain an event semantic vector; A review logic database is constructed based on the event structured data and the event semantic vector.

3. The process compliance review method according to claim 2, characterized in that: Building a review logic database based on the event structured data and the event semantic vector specifically includes: Intelligently analyzing the event structured data and the event semantic vector through the pre-trained AI model to obtain an intelligent analysis result; Performing data processing on the intelligent analysis results to obtain a processed intelligent analysis database; The output format of the processed intelligent analysis database is controlled through a standardized prompting project to obtain a review logic database.

4. The process compliance review method according to claim 3 is characterized in that: Performing data processing on the intelligent analysis results to obtain a processed intelligent analysis database specifically includes: Perform event relationship verification based on the intelligent analysis result to obtain a verified intelligent analysis data set; Completing the relationship description based on the intelligent parsing result to obtain a completed intelligent parsing data set; The verified intelligent analysis data set and the completed intelligent analysis data set are integrated to obtain a processed intelligent analysis database.

5. The process compliance review method according to any one of claims 1 to 4, characterized in that: Based on the review logic database, the processed text data to be reviewed is subjected to compliance review and detection to obtain corresponding process compliance review results, which specifically include: Based on the review logic database, the processed text data to be reviewed is mapped and compared to obtain a data mapping comparison result; Based on dynamic rule data, the data mapping comparison results are checked for violations to obtain corresponding process compliance review results.

6. The process compliance review method according to claim 5, characterized in that: Based on dynamic rule data, the data mapping comparison results are tested for violations and the corresponding process compliance review results are obtained, including: Perform violation detection on data mapping comparison results based on dynamic rule data; If the violation detection is qualified, the normal process compliance review result is obtained; If the violation detection is unqualified, the abnormal process compliance review result is obtained and the corresponding violation review report is output.

7. The process compliance review method according to claim 5, characterized in that: The dynamic rule data generation method includes manual setting and input and automatic generation of the pre-trained AI model based on process review rules.

8. A process compliance review device, characterized in that: The process compliance review device comprises: Data processing module: performs data preprocessing on the text data to be reviewed to obtain the processed text data to be reviewed; Database construction module: constructing a review logic database based on the pre-trained AI model and the processed text data to be reviewed; Process review module: Based on the review logic database, the processed text data to be reviewed is subjected to compliance review and detection to obtain corresponding process compliance review results.

9. A process compliance review device, characterized in that: The process compliance review device comprises: a memory, a processor, and a process compliance review program stored in the memory and executable on the processor, wherein the process compliance review program is configured to implement the process compliance review method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, it can implement the steps in the process compliance review method described in any one of claims 1 to 7.

Images