Financial reimbursement intelligent auditing system and method based on large model

Abstract

The invention discloses a financial reimbursement intelligent auditing system and method based on a large model, and relates to the technical field of informatization. Comprising the following steps: step 1, obtaining audit data through a front-end interface: obtaining an attachment to be audited carrying a bill number, form information and a flow name; 2, extracting attachment information; 3, matching an auditing rule and executing the steps of presetting the auditing rule which covers invoice types, invoice expiration, invoice details and information whether attachment uploading is complete or not; matching preset auditing rules according to the cost items, performing auditing one by one according to the form information and the attachment information, and checking whether the reimbursement receipts meet various rule requirements or not; and 4, generating and storing an auditing report: after the auditing is completed, pushing the business document for manual reauditing, automatically generating the auditing report, the report content comprising problems found in the auditing process, specific conditions conforming to or not conforming to the rules, and translated Chinese file link information, and the auditing result of each rule is displayed in the form of a table.

Description

Technical Field

[0001] This invention discloses an intelligent financial reimbursement review system and method based on a large model, which relates to the field of information technology. Background Technology

[0002] In current corporate finance operations, expense reimbursement review is an important yet tedious task. Currently, expense reimbursement reviews primarily rely on manual processing. Reviewers need to meticulously check each expense report's form information and attachments, verifying invoice type compliance, expiration dates, invoice details completeness, and the integrity of uploaded attachments, among other things. This manual review method has several problems. First, it's inefficient, especially with high volumes of expense reports; a large backlog can hinder the approval process. Second, manual review is susceptible to subjective biases and fatigue among reviewers, making accuracy difficult to guarantee and potentially leading to omissions or errors, thus posing financial risks to the company.

[0003] Although some companies have already introduced information systems to assist in the review of financial reimbursements, most existing systems have limited functions and fixed review rules, making it difficult to flexibly adjust review standards according to changes in business operations. Furthermore, their level of intelligence in areas such as extracting attachment information and matching and executing review rules is low, still requiring a significant amount of manual intervention. As a result, they cannot fully meet companies' needs for efficient, accurate, and intelligent financial reimbursement reviews. Summary of the Invention

[0004] This invention addresses the problems of existing technologies by providing an intelligent financial reimbursement review system and method based on a large model. It solves the problems of low efficiency and poor accuracy of existing manual review, as well as the low level of intelligence and inflexible review rules of existing information-based review systems. This improves the efficiency and accuracy of financial reimbursement review, reduces manual intervention, and provides strong support for financial reimbursement approval work.

[0005] The specific solution proposed in this invention is as follows: This invention provides a method for intelligent auditing of financial reimbursements based on a large model, comprising: Step 1: Obtain review data through the front-end interface: Obtain the attachment to be reviewed, which contains the document number, form information, and process name; Step 2: Extract attachment information: Query the attachment information saved during the form filling stage from the database based on the document number. After the attachment information is found, compare it with the attachment list. If there are attachments added after the form is filled, extract the attachment content. If the attachment information is not found, call the invoice information extraction interface and attachment information extraction interface in the automated form filling to extract information from all attachments. The extracted information includes invoice details, attachment list, and purchase list. Step 3: Match and execute audit rules: Preset audit rules, covering invoice type, invoice expiration, invoice details, and whether the uploaded attachments are complete; match the pre-set audit rules according to the expense items, and audit each item according to the form information and attachment information to check whether the reimbursement documents meet the requirements of each rule; Step 4: Generate and save the audit report: After the audit is completed, the business documents will be sent to manual review, and an audit report will be automatically generated. The report includes the problems found during the audit process, the specific circumstances of compliance or non-compliance with the rules, and the link information of the translated Chinese version of the document. The audit results of each rule are displayed in tabular form.

[0006] Furthermore, in step 2 of the intelligent auditing method for financial reimbursement based on a large model, the audit information details required for auditing are extracted from the form information and attachments. The input items are the document number, form information, attachments under the process name and attachments in the form, and the output item is the audit information details. If the audit information extraction fails, the extracted audit information details are returned. If the audit information extraction fails, the specific reason for the failure is returned.

[0007] Furthermore, in step 3 of the intelligent financial reimbursement review method based on a large model, the review rules are dynamically loaded through a knowledge base, and the review standards are adjusted in real time to ensure the compliance and accuracy of the review results. The review rule query interface is used to match the corresponding review rules for the expense item type. The input item is the expense item type, and the output item is the list of review rules. If the rule matching fails, the list of matched review rules is returned. If the rule matching fails, it returns that no review rule has been set for this expense item type.

[0008] Furthermore, in step 3 of the aforementioned intelligent auditing method for financial reimbursement based on a large model, an intelligent auditing interface is used to obtain a list of auditing rules from the auditing rule management interface according to the expense item type, and to obtain auditing information details from the auditing information extraction interface. Each auditing rule is analyzed and audited one by one to identify compliant and non-compliant items, and the auditing result data is returned. The input items are the expense item type and auditing information details, and the output items are the auditing results of each auditing rule. If the audit is successful, a complete auditing result is returned; if the audit fails, the specific reason for the audit failure is returned.

[0009] Furthermore, in step 4 of the intelligent financial reimbursement review method based on a large model, the review report generation interface is used to perform statistical analysis on the review results of each rule, generate a list of review results and detailed review details, and form a review report. The input items are the review results of each rule, and the output item is the review report. If the review report generation fails, the generated review report file is returned. If the review report generation fails, the specific reason is returned.

[0010] This invention also provides a financial reimbursement intelligent auditing system based on a large model, including a display module, an audit information extraction module, an audit rule management module, an intelligent auditing module, and an audit report generation module. The display module retrieves review data through the front-end interface: it obtains attachments to be reviewed, containing document numbers, form information, and process names; and it displays the overall intelligent financial reimbursement review process accordingly. The audit information extraction module extracts attachment information. Based on the document number, it queries the database for attachment information saved during the form filling stage. After finding the attachment information, it compares it with the attachment list. If there are attachments added after the form is filled out, the attachment content is extracted. If the attachment information is not found, the invoice information extraction interface and attachment information extraction interface in the automated form filling are called to extract information from all attachments. The extracted information includes detailed invoice information, attachment list, and purchase list. The intelligent review module matches and executes review rules. The review rule management module presets review rules, covering invoice type, invoice expiration, invoice details, and whether the uploaded attachments are complete. The intelligent review module matches the pre-set review rules according to the expense items and reviews each form and attachment item to check whether the reimbursement documents meet the requirements of each rule. The audit report generation module generates and saves audit reports: After the audit is completed, the business documents are pushed to manual review, and an audit report is automatically generated. The report content includes the problems found during the audit process, the specific circumstances of compliance or non-compliance with the rules, and the link information of the translated Chinese version of the document. The audit results of each rule are displayed in the form of a table.

[0011] Furthermore, the audit information extraction module of the intelligent financial reimbursement audit system based on a large model extracts the audit information details required for auditing from form information and attachments. The input items are document number, form information, attachments under process name and attachments in the form, and the output item is audit information details. If the audit information extraction fails, the extracted audit information details are returned. If the audit information extraction fails, the specific reason for the failure is returned.

[0012] Furthermore, the audit rule management module of the intelligent financial reimbursement audit system based on a large model dynamically loads audit rules through a knowledge base and adjusts audit standards in real time to ensure the compliance and accuracy of audit results. The intelligent audit module uses the audit rule query interface to match the corresponding audit rules for the expense item type. The input is the expense item type, and the output is a list of audit rules. If the rule matching fails, it returns a list of matched audit rules. If the rule matching fails, it returns that no audit rules have been set for this expense item type.

[0013] Furthermore, the intelligent audit module of the financial reimbursement intelligent audit system based on a large model utilizes an intelligent audit interface to obtain a list of audit rules from the audit rule management interface based on the expense item type, and obtain audit information details from the audit information extraction interface. It analyzes and audits each audit rule one by one, identifies compliant and non-compliant items, and returns audit result data. The input items are the expense item type and audit information details, and the output items are the audit results of each audit rule. If the audit is successful, a complete audit result is returned; if the audit fails, the specific reason for the failure is returned.

[0014] Furthermore, the audit report generation module of the intelligent financial reimbursement audit system based on a large model utilizes the audit report generation interface to perform statistical analysis on the audit results of various rules, generate an audit result and a detailed audit details list, and form an audit report. The input items are the audit results of various rules, and the output item is the audit report. If the audit report generation fails, the generated audit report file is returned. If the audit report generation fails, the specific reason is returned.

[0015] The advantages of this invention are: (1) Improve audit efficiency: By automating the acquisition of audit data, extraction of attachment information, matching and execution of audit rules and generation of audit reports, a large amount of traditional manual operations are replaced, which greatly reduces the workload of auditors, shortens the audit time of financial reimbursement, improves audit efficiency, and effectively solves the problems of reimbursement form accumulation and slow approval progress.

[0016] (2) Improve the accuracy of audit: With the help of big data model technology, the audit rules are accurately matched and strictly implemented, avoiding omissions and errors caused by subjective factors and fatigue during the manual audit process, thus improving the accuracy of financial reimbursement audit and reducing corporate financial risks.

[0017] (3) Flexible and adjustable audit rules: The audit rules are dynamically loaded through the knowledge base of the audit rule management module. The audit standards can be adjusted in real time according to changes in enterprise business and policy adjustments, so as to meet the audit needs of enterprises at different stages and in different business scenarios, and enhance the adaptability and flexibility of the system.

[0018] (4) Assisting manual review: The generated audit report records key information in the audit process in detail, including issues, compliance status, etc., and displays them in an intuitive table format. It also supports viewing details, providing sufficient reference for manual review and helping to improve the efficiency and quality of manual review. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the application process of the method of the present invention. Detailed Implementation

[0020] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, so that those skilled in the art can better understand and implement the present invention. However, the embodiments described are not intended to limit the present invention. Example

[0021] This invention provides a method for intelligent auditing of financial reimbursements based on a large model, comprising: Step 1: Obtain the review data through the front-end interface: Obtain the attachment to be reviewed, which contains the document number, form information, and process name.

[0022] Step 2: Extract Attachment Information: Extract key information from form information and attachments, including invoice details, attachment list, purchase list, etc. From the form information, extract invoice details and attachment name list. For attachment information, first query the database based on the document number to retrieve attachment information saved during the form filling stage. After retrieving the attachment information, compare it with the attachment list transmitted by the reimbursement system. If there are attachments added after filling in the form, their content needs to be extracted. If the attachment information is not found, call the invoice information extraction and attachment information extraction interfaces in the automated form filling to extract information from all attachments. This module does not provide page operations and has an audit information extraction interface. Its function is to extract the audit information details required for auditing from form information and attachments. The input items are document number, form information, attachments under the process name, and attachments in the form. The output item is the audit information details. If successful, the extracted audit information details are returned; if the audit information extraction fails, the specific reason for the failure is returned.

[0023] Step 3: Match and Execute Audit Rules: This step involves matching audit rules based on expense item type, covering multiple audit points such as invoice type, invoice expiration, invoice details, and completeness of uploaded attachments. A knowledge base is used to dynamically load audit rules, allowing for real-time adjustments to audit standards and ensuring the compliance and accuracy of audit results. The audit rule query interface matches expense item types with corresponding audit rules. The input is the expense item type, and the output is a list of audit rules. A list of matched audit rules is returned upon successful matching; otherwise, "No audit rules have been set for this expense item type" is returned.

[0024] The system matches pre-defined audit rules to expense items. These rules cover invoice type, invoice expiration, invoice details, and the completeness of uploaded attachments. Each rule is reviewed based on form and attachment information to ensure compliance with all requirements. Utilizing an intelligent audit interface, the system retrieves a list of audit rules from the audit rule management interface and detailed audit information from the audit information extraction interface. Each rule is analyzed and audited individually to identify compliance and non-compliance items, and the audit results are returned. Inputs include expense item type and audit information details; outputs are the audit results for each rule. A complete audit result is returned if all results are received; otherwise, the specific reason for the failure is returned.

[0025] Step 4: Generate and Save the Audit Report: After the audit is completed, the business documents are sent for manual review, and an audit report is automatically generated. The report includes issues found during the audit process, specific details of compliance or non-compliance with rules, and links to translated Chinese versions of the documents. The audit results for each rule are displayed in tabular form. The audit report is generated based on the audit results for each rule, detailing issues found during the audit process, specific details of compliance or non-compliance with rules, etc. The audit results for each rule are displayed in tabular form to assist manual review in document auditing. Using the audit report generation interface, statistical analysis is performed on the audit results for each rule, generating an audit result and detailed audit details list, forming the audit report. The input is the audit results for each rule, and the output is the audit report. If the audit report generation fails, the generated audit report file is returned; if the audit report generation fails, the specific reason is returned. Example

[0026] This invention also provides a financial reimbursement intelligent auditing system based on a large model, including a display module, an audit information extraction module, an audit rule management module, an intelligent auditing module, and an audit report generation module. The display module obtains audit data through the front-end interface: it retrieves attachments to be audited, carrying document numbers, form information, and process names; and it displays the overall intelligent audit process for financial reimbursement. For example, it receives document numbers, form information, attachments under the process name, and attachments in the form from the reimbursement system. After processing through audit information extraction, intelligent auditing, and audit report generation, the intelligently generated audit report is returned to the reimbursement system for storage as a file download link. The audit report is also displayed in the lower right corner of the initial review page of the reimbursement system, making it convenient for finance personnel and relevant approvers to view, providing an intuitive reference for reimbursement approval. In terms of interface design, the audit report is displayed in the lower right corner of the initial review page of the reimbursement system. Clicking "View Details" expands to a new webpage displaying the audit report details. This module also has a general expense report review interface. The interface function is to receive general expense report form information and attachments, review them according to preset rules, and return a review report. The input items are document number, form information, attachments under process name and attachments in the form. The output item is a link to download the review report. Under normal circumstances, the review report download link is returned. If the review fails, "General expense report review failed" and the reason for the failure are returned, and an exception log is recorded.

[0027] The audit information extraction module extracts attachment information. Based on the document number, it queries the database for attachment information saved during the form filling stage. After finding the attachment information, it compares it with the attachment list. If there are attachments added after the form is filled out, the attachment content is extracted. If the attachment information is not found, the invoice information extraction interface and attachment information extraction interface in the automated form filling are called to extract information from all attachments. The extracted information includes detailed invoice information, attachment list, and purchase list. The intelligent review module matches and executes review rules. The review rule management module presets review rules, covering invoice type, invoice expiration, invoice details, and whether the uploaded attachments are complete. The intelligent review module matches the pre-set review rules according to the expense items and reviews each form and attachment item to check whether the reimbursement documents meet the requirements of each rule. The audit report generation module generates and saves audit reports: After the audit is completed, the business documents are pushed to manual review, and an audit report is automatically generated. The report content includes the problems found during the audit process, the specific circumstances of compliance or non-compliance with the rules, and the link information of the translated Chinese version of the document. The audit results of each rule are displayed in the form of a table.

[0028] The information interaction and execution process between the modules in the above system are based on the same concept as the method embodiment of the present invention, and the specific details can be found in the description in the method embodiment of the present invention, and will not be repeated here.

[0029] Similarly, the advantages of the system of the present invention are: (1) Improve audit efficiency: By automating the acquisition of audit data, extraction of attachment information, matching and execution of audit rules and generation of audit reports, a large amount of traditional manual operations are replaced, which greatly reduces the workload of auditors, shortens the audit time of financial reimbursement, improves audit efficiency, and effectively solves the problems of reimbursement form accumulation and slow approval progress.

[0030] (2) Improve the accuracy of audit: With the help of big data model technology, the audit rules are accurately matched and strictly implemented, avoiding omissions and errors caused by subjective factors and fatigue during the manual audit process, thus improving the accuracy of financial reimbursement audit and reducing corporate financial risks.

[0031] (3) Flexible and adjustable audit rules: The audit rules are dynamically loaded through the knowledge base of the audit rule management module. The audit standards can be adjusted in real time according to changes in enterprise business and policy adjustments, so as to meet the audit needs of enterprises at different stages and in different business scenarios, and enhance the adaptability and flexibility of the system.

[0032] (4) Assisting manual review: The generated audit report records key information in the audit process in detail, including issues, compliance status, etc., and displays them in an intuitive table format. It also supports viewing details, providing sufficient reference for manual review and helping to improve the efficiency and quality of manual review.

[0033] It should be noted that not all steps and modules in the above processes and system structures are mandatory; some steps or modules can be omitted as needed. The execution order of each step is not fixed and can be adjusted as required. The system structures described in the above embodiments can be physical or logical structures. That is, some modules may be implemented by the same physical entity, or some modules may be implemented by multiple physical entities, or they may be jointly implemented by certain components in multiple independent devices.

[0034] The above-described embodiments are merely preferred embodiments provided to fully illustrate the present invention, and the scope of protection of the present invention is not limited thereto. Equivalent substitutions or modifications made by those skilled in the art based on the present invention are all within the scope of protection of the present invention. The scope of protection of the present invention is defined by the claims.

Claims

1. A method for intelligent auditing of financial reimbursements based on a large model, characterized by: include: Step 1: Obtain review data through the front-end interface: Obtain the attachment to be reviewed, which contains the document number, form information, and process name; Step 2: Extract attachment information: Query the attachment information saved during the form filling stage from the database based on the document number. After the attachment information is found, compare it with the attachment list. If there are attachments added after the form is filled, extract the attachment content. If the attachment information is not found, call the invoice information extraction interface and attachment information extraction interface in the automated form filling to extract information from all attachments. The extracted information includes invoice details, attachment list, and purchase list. Step 3: Match and execute audit rules: Preset audit rules, covering invoice type, invoice expiration, invoice details, and whether the uploaded attachments are complete; match the pre-set audit rules according to the expense items, and audit each item according to the form information and attachment information to check whether the reimbursement documents meet the requirements of each rule; Step 4: Generate and save the audit report: After the audit is completed, the business documents will be sent to manual review, and an audit report will be automatically generated. The report includes the problems found during the audit process, the specific circumstances of compliance or non-compliance with the rules, and the link information of the translated Chinese version of the document. The audit results of each rule are displayed in tabular form.

2. The intelligent auditing method for financial reimbursement based on a large model as described in claim 1 is characterized in that, in step 2, the audit information details required for auditing are extracted from the form information and attachments. The input items are the document number, form information, attachments under the process name and attachments in the form, and the output item is the audit information details. If the audit information extraction fails, the extracted audit information details are returned. If the audit information extraction fails, the specific reason for the failure is returned.

3. The intelligent auditing method for financial reimbursement based on a large model as described in claim 1, characterized in that: In step 3, the audit rules are dynamically loaded through the knowledge base, and the audit standards are adjusted in real time to ensure the compliance and accuracy of the audit results. The audit rule query interface is used to match the corresponding audit rules for the type of expense item. The input item is the expense item type, and the output item is the list of audit rules. If a rule match is successful, a list of matching review rules will be returned. If a rule match fails, it will return that no review rules have been set for this fee item type.

4. The intelligent auditing method for financial reimbursement based on a large model as described in claim 1, characterized in that: In step 3, the intelligent audit interface is used to obtain the list of audit rules from the audit rule management interface according to the type of expense item, and the audit information details from the audit information extraction interface. Each audit rule is analyzed and audited one by one to identify compliant and non-compliant items, and the audit result data is returned. The input items are the expense item type and audit information details, and the output items are the audit results of each audit rule. If the audit is successful, a complete audit result is returned. If the audit fails, the specific reason for the audit failure is returned.

5. The intelligent auditing method for financial reimbursement based on a large model as described in claim 1, characterized in that: In step 4, the audit report generation interface is used to perform statistical analysis on the audit results of each rule, generate an audit result and a detailed audit details list, and form an audit report. The input is the audit results of each rule, and the output is the audit report. If the audit report generation fails, the generated audit report file is returned. If the audit report generation fails, the specific reason is returned.

6. A financial reimbursement intelligent approval system based on a large model, characterized by: It includes a display module, an audit information extraction module, an audit rule management module, an intelligent audit module, and an audit report generation module. The display module retrieves review data through the front-end interface: it obtains attachments to be reviewed, containing document numbers, form information, and process names; and it displays the overall intelligent financial reimbursement review process accordingly. The audit information extraction module extracts attachment information. Based on the document number, it queries the database for attachment information saved during the form filling stage. After finding the attachment information, it compares it with the attachment list. If there are attachments added after the form is filled out, the attachment content is extracted. If the attachment information is not found, the invoice information extraction interface and attachment information extraction interface in the automated form filling are called to extract information from all attachments. The extracted information includes detailed invoice information, attachment list, and purchase list. The intelligent review module matches and executes review rules. The review rule management module presets review rules, covering invoice type, invoice expiration, invoice details, and whether the uploaded attachments are complete. The intelligent review module matches the pre-set review rules according to the expense items and reviews each form and attachment item to check whether the reimbursement documents meet the requirements of each rule. The audit report generation module generates and saves audit reports: After the audit is completed, the business documents are pushed to manual review, and an audit report is automatically generated. The report content includes the problems found during the audit process, the specific circumstances of compliance or non-compliance with the rules, and the link information of the translated Chinese version of the document. The audit results of each rule are displayed in the form of a table.

7. The intelligent financial reimbursement review system based on a large model according to claim 6, characterized in that: The audit information extraction module extracts the audit information details required for auditing from form information and attachments. The input items are document number, form information, attachments under process name and attachments in the form, and the output item is audit information details. If the audit information extraction fails, the module returns the extracted audit information details. If the audit information extraction fails, the module returns the specific reason for the failure.

8. The intelligent financial reimbursement review system based on a large model according to claim 6, characterized in that: The audit rule management module dynamically loads audit rules through a knowledge base and adjusts audit standards in real time to ensure the compliance and accuracy of audit results. The intelligent audit module uses the audit rule query interface to match the corresponding audit rules for the types of expense items. The input is the expense item type, and the output is a list of audit rules. If a rule match is successful, a list of matching review rules will be returned. If a rule match fails, it will return that no review rules have been set for this fee item type.

9. The intelligent financial reimbursement review system based on a large model according to claim 6, characterized in that: The intelligent audit module utilizes the intelligent audit interface to retrieve a list of audit rules from the audit rule management interface based on the expense item type, and obtain audit information details from the audit information extraction interface. It analyzes and audits each audit rule one by one, identifies compliant and non-compliant items, and returns audit result data. The input items are the expense item type and audit information details, and the output items are the audit results of each audit rule. If the audit is successful, it returns a complete audit result; if the audit fails, it returns the specific reason for the failure.

10. A financial reimbursement intelligent review system based on a large model as described in claim 6, characterized in that the review... The report generation module uses the audit report generation interface to perform statistical analysis on the audit results of various rules, generate an audit result and a detailed audit details list, and form an audit report. The input is the audit results of various rules, and the output is the audit report. If the audit report generation fails, the module returns the generated audit report file. If the audit report generation fails, the module returns the specific reason.

Images