Enterprise investigation material quality inspection method and device, storage medium and program product

Abstract

The invention discloses an enterprise investigation material quality inspection method and device, a storage medium and a program product. The method comprises the following steps: under the condition that a target enterprise submits a business application to a financial institution, obtaining a full-duty investigation material of the target enterprise; determining N quality inspection items based on the business application, N being a positive integer; based on the N quality inspection items, performing quality inspection on the full-duty investigation material by using a large language model to obtain a target quality inspection result; and determining a processing mode of the business application based on the target quality inspection result. According to the method and the device, the problem of low quality inspection efficiency caused by manual quality inspection of the full-duty investigation material in the process of handling enterprise business in the prior art is solved.

Description

Technical Field

[0001] This application relates to the fields of financial technology and artificial intelligence, and more specifically, to a method, apparatus, storage medium, and program product for quality inspection of corporate investigation materials. Background Technology

[0002] Financial institution branches receive customer business applications and due diligence materials (hereinafter referred to as due diligence materials). Business personnel manually enter information, check data in credit reporting, business registration and other systems to verify data, and review due diligence materials such as multi-angle photos taken on-site by the enterprise and videos of the customer's business intentions recorded by the customer one by one to verify the authenticity, completeness and compliance of the materials, and comprehensively judge the potential risk points of the enterprise.

[0003] However, due to the diverse sources and complex formats of due diligence materials, traditional audits require manual comparison, key information retrieval, and verification processes, which are inefficient, cumbersome, and time-consuming. Furthermore, human error can inevitably lead to omissions of key information, data entry errors, and inaccurate verification. On the other hand, verifying the authenticity of due diligence materials relies on manual visual inspection and experience-based judgment, which has limitations and is difficult to standardize. The ability to systematically identify forged or altered materials is significantly weak, creating hidden risks for illegal transactions.

[0004] Therefore, manual review of due diligence materials in related technologies suffers from low efficiency and high risk. When faced with massive amounts of information and complex risk scenarios, its accuracy, comprehensiveness, and risk insight are all insufficient, making it difficult to support the intelligent risk management needs of financial institutions.

[0005] There is currently no effective solution to the problem of low efficiency in manually inspecting due diligence materials during the business processing of related technologies. Summary of the Invention

[0006] The main purpose of this application is to provide a method, device, storage medium, and program product for quality inspection of enterprise investigation materials, so as to solve the problem of low quality inspection efficiency when manually inspecting due diligence materials during the handling of enterprise business.

[0007] To achieve the above objectives, according to one aspect of this application, a method for quality inspection of enterprise due diligence materials is provided. The method includes: obtaining the due diligence materials of a target enterprise when it submits a business application to a financial institution; determining N quality inspection items based on the business application, where N is a positive integer; performing quality inspection on the due diligence materials using a large language model based on the N quality inspection items to obtain a target quality inspection result; and determining the processing method for the business application based on the target quality inspection result.

[0008] Further, based on N quality inspection items, a large language model is used to perform quality inspection on the due diligence materials to obtain a target quality inspection result, including: determining N quality inspection tasks based on the N quality inspection items and the due diligence materials, wherein each quality inspection task is used to perform quality inspection on the materials to be inspected by the large language model, and each quality inspection task includes: configuration parameters associated with one of the quality inspection items and the materials to be inspected by the quality inspection item; executing each quality inspection task and polling the task execution status and quality inspection result of each quality inspection task to obtain a target detection result; and determining the target quality inspection result based on the target detection result.

[0009] Further, polling the task execution status and quality inspection results of each quality inspection task to obtain the target detection result includes: polling to check whether the task execution status of each quality inspection task is completed to obtain an initial detection result; if the initial detection result indicates that the task execution status of all quality inspection tasks is completed, obtaining the quality inspection results generated by all quality inspection tasks to obtain the target detection result; if the initial detection result indicates that the task execution status of any quality inspection task is in progress, obtaining the task execution time of that quality inspection task; if the task execution time of that quality inspection task exceeds a preset time threshold, terminating the quality inspection task and generating an anomaly diagnosis log for that quality inspection task; and obtaining the target detection result based on the anomaly diagnosis log and all generated quality inspection results.

[0010] Furthermore, the due diligence materials include: textual materials and video materials. The video materials include: videos representing the target company's business intentions and on-site images of the target company. Each quality inspection task is executed, including: determining the quality inspection service to be invoked for each quality inspection task. The types of quality inspection services include: information quality inspection service, video quality inspection service, and image quality inspection service. The information quality inspection service uses the large language model to detect whether the textual materials are authentic; the video quality inspection service uses the large language model to detect whether the business intentions expressed in the video match the business application; and the image quality inspection service uses the large language model to detect whether the target company's business premises are authentic. Each quality inspection task invokes the associated quality inspection service and uses the large language model to perform quality inspection on the due diligence materials.

[0011] Furthermore, after determining N quality inspection tasks based on the N quality inspection items and the due diligence materials, the method further includes: adding the N quality inspection tasks to a target message queue; and executing each quality inspection task, including: reading each quality inspection task from the target message queue and executing the read quality inspection task.

[0012] Furthermore, after performing quality inspection on the due diligence materials based on N quality inspection items using a large language model to obtain the target quality inspection result, the method further includes: based on the target quality inspection result, marking the quality inspection result of each of the quality inspection items to obtain N marked quality inspection items; visually displaying the N marked quality inspection items, and receiving feedback information from the target object regarding the N marked quality inspection items, wherein the target object includes: the object in the financial institution that processes the business application for the target enterprise, and the feedback information is used to indicate whether the quality inspection result of each quality inspection item is accurate.

[0013] Further, based on the business application, N quality inspection items are determined, including: generating a quality inspection list based on the business application, wherein the quality inspection list includes M quality inspection items, where M is an integer greater than N; displaying the quality inspection list on a target interface, and using the target interface to receive the N quality inspection items selected by the target object, wherein the target interface includes at least M of the quality inspection items and check buttons for the M quality inspection items.

[0014] To achieve the above objectives, according to another aspect of this application, a device for quality inspection of enterprise investigation materials is provided. The device includes: an acquisition unit for acquiring due diligence materials of a target enterprise when the target enterprise submits a business application to a financial institution; a first determining unit for determining N quality inspection items based on the business application, where N is a positive integer; a quality inspection unit for performing quality inspection on the due diligence materials using a large language model based on the N quality inspection items to obtain a target quality inspection result; and a second determining unit for determining the processing method of the business application based on the target quality inspection result.

[0015] Further, the quality inspection unit includes: a first determining subunit, configured to determine N quality inspection tasks based on N quality inspection items and the due diligence materials, wherein each quality inspection task is used to perform quality inspection on the materials to be inspected by the large language model, and each quality inspection task includes: configuration parameters associated with one of the quality inspection items and the materials to be inspected by the quality inspection item; a first processing subunit, configured to execute each quality inspection task and poll and detect the task execution status and quality inspection results of each quality inspection task to obtain a target detection result; and a second determining subunit, configured to determine the target quality inspection result based on the target detection result.

[0016] Further, the first processing subunit includes: a detection module, used to poll and detect whether the task execution status of each quality inspection task is completed, and obtain an initial detection result; an acquisition module, used to acquire the quality inspection results generated by all the quality inspection tasks when the initial detection result indicates that the task execution status of all the quality inspection tasks is completed, and obtain the target detection result; and a first processing module, used to acquire the task execution time of any quality inspection task when the initial detection result indicates that the task execution status of any quality inspection task is in progress, terminate the quality inspection task when the task execution time of the quality inspection task exceeds a preset time threshold, generate an anomaly diagnosis log for the quality inspection task, and obtain the target detection result based on the anomaly diagnosis log and all generated quality inspection results.

[0017] Furthermore, the due diligence materials include: textual materials and image data. The image data includes: videos representing the target company's business intentions and on-site images of the target company. The processing subunit further includes: a determination module, used to determine the quality inspection service to be invoked for each quality inspection task, wherein the types of quality inspection services include: information quality inspection service, video quality inspection service, and image quality inspection service. The information quality inspection service is used to detect whether the textual materials are authentic using the large language model. The video quality inspection service is used to detect whether the business intentions expressed in the video are consistent with the business application using the large language model. The image quality inspection service is used to detect whether the target company's business premises are authentic using the large language model. The second processing module is used to invoke the quality inspection service associated with each quality inspection task and to perform quality inspection on the due diligence materials using the large language model.

[0018] Furthermore, the enterprise investigation material quality inspection device includes: an adding unit, used to add N quality inspection tasks to a target message queue after determining N quality inspection tasks based on N quality inspection items and the due diligence materials; and a processing subunit including: an execution module, used to read each quality inspection task from the target message queue and execute the read quality inspection task.

[0019] Furthermore, the enterprise investigation material quality inspection device also includes: a marking unit, used to perform quality inspection on the due diligence materials based on N quality inspection items using a large language model, and after obtaining the target quality inspection result, mark the quality inspection result of each of the quality inspection items based on the target quality inspection result, to obtain N marked quality inspection items; and a processing unit, used to visualize the N marked quality inspection items and receive feedback information from the target object on the N marked quality inspection items, wherein the target object includes: the object in the financial institution that processes the business application for the target enterprise, and the feedback information is used to indicate whether the quality inspection result of each quality inspection item is accurate.

[0020] Further, the first determining unit includes: a generation subunit, used to generate a quality inspection list based on the business application, wherein the quality inspection list includes M quality inspection items, where M is an integer greater than N; and a second processing subunit, used to display the quality inspection list on a target interface, and to receive N quality inspection items selected by the target object using the target interface, wherein the target interface includes at least M quality inspection items and check buttons for the M quality inspection items.

[0021] According to another aspect of this application, a computer-readable storage medium is provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is executed, it controls the device where the computer-readable storage medium is located to perform the enterprise survey material quality inspection method.

[0022] According to another aspect of this application, an electronic device is provided, comprising: a memory storing an executable program; and a processor for running the program, wherein the program, when running, executes the enterprise survey material quality inspection method.

[0023] According to another aspect of this application, a computer program product is provided, including computer instructions that, when executed by a processor, implement the steps of the enterprise survey material quality inspection method.

[0024] This application employs the following method: When a target company submits a business application to a financial institution, its due diligence materials are obtained; based on the business application, N quality inspection items are determined, where N is a positive integer; based on these N quality inspection items, a large language model is used to inspect the due diligence materials, yielding a target quality inspection result; based on the target quality inspection result, the processing method for the business application is determined. This solves the technical problem of low efficiency in manual quality inspection of due diligence materials during the processing of business transactions in related technologies. In this application, by setting multiple quality inspection items and using a large language model to inspect the due diligence materials, the low efficiency and high error rate of manual quality inspection of due diligence materials in related technologies are avoided, thereby achieving the technical effect of improving the efficiency and accuracy of due diligence material quality inspection. Attached Figure Description

[0025] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings:

[0026] Figure 1 A hardware structure block diagram of a computer terminal for implementing a quality inspection method for enterprise survey materials is shown.

[0027] Figure 2 This is a flowchart of a method for quality inspection of enterprise investigation materials provided in the embodiments of this application;

[0028] Figure 3 This is a schematic diagram of an enterprise investigation material quality inspection system provided according to an embodiment of this application;

[0029] Figure 4 This is a flowchart of the quality inspection process for enterprise investigation materials provided in the embodiments of this application;

[0030] Figure 5 This is a schematic diagram of a quality inspection device for enterprise investigation materials provided in the embodiments of this application;

[0031] Figure 6 This is a structural block diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0032] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0033] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0034] It should be noted that the enterprise investigation material quality inspection method and apparatus in this application can be used in the fintech field for quality inspection of due diligence materials, and can also be used in any field other than fintech for quality inspection of due diligence materials. This application does not limit the application field of the enterprise investigation material quality inspection method and apparatus.

[0035] For ease of description, some terms or nouns involved in the various embodiments of the present invention will be explained below.

[0036] Customer due diligence: When establishing a business relationship with a customer, conducting a one-time transaction exceeding a specified amount, and during the duration of the business relationship, financial institutions identify, verify, register, and retain customer identity information, and take appropriate due diligence measures based on the risk situation to understand the purpose and nature of the customer's establishment of the business relationship and transactions.

[0037] It should be noted that the information collected in this application (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for display, data used for analysis, etc.) are information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of this data all comply with relevant laws, regulations, and standards, necessary confidentiality measures have been taken, and they do not violate public order and good morals. Corresponding access points are provided for users to choose to authorize or refuse. For example, interfaces are set up between this system and relevant users or organizations, providing users with corresponding access points to choose to agree to or refuse automated decision-making results; if the user chooses to refuse, the process proceeds to the expert decision-making stage.

[0038] Example 1

[0039] According to an embodiment of this application, a method embodiment for quality inspection of enterprise investigation materials is also provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0040] The method embodiment provided in Embodiment 1 of this application can be executed on a mobile terminal, computer terminal, or similar computing device. Figure 1 A hardware block diagram of a computer terminal (or mobile device) for implementing quality inspection methods for enterprise survey materials is shown. Figure 1 As shown, the computer terminal 10 (or mobile device) may include one or more processors 102 (shown as 102a, 102b, ..., 102n in the figure) 102 (processor 102 may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of a BUS bus), a network interface, a power supply, and / or a camera. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0041] It should be noted that the aforementioned one or more processors 102 and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the computer terminal 10 (or mobile device). As involved in the embodiments of this application, the data processing circuits serve as a processor control mechanism (e.g., selection of a variable resistor termination path connected to an interface).

[0042] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the enterprise investigation material quality inspection method in this embodiment. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the aforementioned enterprise investigation material quality inspection method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0043] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0044] The display can be, for example, a touchscreen liquid crystal display (LCD), which allows the user to interact with the user interface of the computer terminal 10 (or mobile device).

[0045] Under the aforementioned operating environment, this application provides the following: Figure 2 The methods for quality inspection of enterprise survey materials are shown. Figure 2 This is a flowchart of the enterprise investigation material quality inspection method according to Embodiment 1 of this application.

[0046] Step S201: Obtain the due diligence materials of the target company when the target company submits a business application to the financial institution.

[0047] The target enterprises mentioned above may include: financial institutions that provide financial services; the business applications mentioned above may include: handling one-time transactions exceeding a specified amount; and may also include: applications for identification, verification, registration, and retention of customer identity information during the business relationship.

[0048] The aforementioned due diligence materials may include, but are not limited to: videos of the client's business intentions recorded by the client, on-site photos of the target company taken by financial institutions during their due diligence visits, and other video materials.

[0049] Step S202: Based on the business application, determine N quality inspection items, where N is a positive integer.

[0050] In this embodiment, business applications can be parsed to identify the type of business application (e.g., account opening, loan application, transfer authorization, etc.), due diligence scenario requirements, and risk concerns. Based on the type of business application, due diligence scenario requirements, and risk concerns, the appropriate quality inspection items can be selected automatically or manually by business personnel from a preset quality inspection item pool based on business rules. These items may include, but are not limited to, the accuracy of customer information, the consistency between the official website address and the address in the due diligence materials, the validity of contact information, the authenticity of video content, and the compliance of images.

[0051] Step S203: Based on N quality inspection items, use a large language model to conduct quality inspection on the due diligence materials and obtain the target quality inspection results.

[0052] For example, the aforementioned large language model can perform in-depth analysis and quality checks on due diligence materials. Through the intelligent analysis of the large language model, problems in due diligence materials can be automatically detected, such as inconsistent address information, vague descriptions, missing evidence, or doubts.

[0053] Step S204: Based on the target quality inspection results, determine the processing method for the business application.

[0054] In this embodiment, the target quality inspection results can be visualized. The target enterprise's personnel in the financial institution (e.g., business personnel) can review the target quality inspection results and determine the processing method for the business application based on the review results, such as whether it is the business corresponding to the business application.

[0055] Through the above steps, this embodiment sets up multiple quality inspection items and utilizes a large language model to inspect due diligence materials. This avoids the low efficiency and high error rate of manual quality inspection of due diligence materials in related technologies, thereby improving the efficiency and accuracy of due diligence material quality inspection. Furthermore, it solves the technical problem of low efficiency in manual quality inspection of due diligence materials during the processing of business transactions in related technologies.

[0056] Optionally, in the enterprise investigation material quality inspection method provided in this application embodiment, based on N quality inspection items, a large language model is used to inspect the due diligence materials to obtain a target quality inspection result. This includes: determining N quality inspection tasks based on the N quality inspection items and the due diligence materials, wherein each quality inspection task is used to inspect the materials inspected by that task using a large language model; each quality inspection task includes: configuration parameters associated with one of the quality inspection items and the materials inspected by that quality inspection item; executing each quality inspection task and polling the task execution status and quality inspection results of each quality inspection task to obtain a target detection result; and determining the target quality inspection result based on the target detection result.

[0057] For example, a series of quality inspection items can be determined based on the type of business application. Each quality inspection item can be used to verify a certain aspect of the due diligence materials, such as the accuracy of customer information or the authenticity of video content. Then, based on the quality inspection items and due diligence materials, corresponding quality inspection tasks can be created. Each task can include the configuration parameters of a quality inspection item and its corresponding materials to be inspected.

[0058] The configuration parameters mentioned above can be used to guide large language models on how to process specific materials. For example, if the quality control item is "video content consistency," the configuration parameters can include keywords to be mentioned in the video, expected speech patterns, etc., so that the model can analyze the video content against these parameters.

[0059] Each quality inspection task can be performed by using a large language model to intelligently analyze the materials associated with that task. For example, the large language model may need to identify and understand the audio content in a video and compare it with expected keywords and phrases; or it may need to analyze an image, identify the text information within it, and match it with information in a database to verify its authenticity.

[0060] During the execution of quality inspection tasks, the execution status of each task can be continuously polled to ensure that the tasks run normally and the inspection results are obtained in a timely manner. If a quality inspection task encounters a fault or timeout during execution, appropriate error handling can be performed, such as retrying, skipping the task, or marking it as an exception.

[0061] Once all quality inspection tasks are completed, the quality inspection results associated with all quality inspection tasks can be summarized to obtain the target quality inspection result and generate a quality inspection result report. It should be noted that the target quality inspection result may include: quality inspection items that have passed the quality inspection, quality inspection items that have not passed the quality inspection, and quality inspection items that have failed due to timeout or malfunction during the execution of the quality inspection task.

[0062] Based on the results of the target inspection, the reviewers can comprehensively evaluate the quality of the entire due diligence material and determine whether it passes the quality inspection. If any material is marked as unqualified, additional review processes can be conducted, or the client can be asked to submit new materials for re-examination.

[0063] By using large language models to intelligently process due diligence materials, the efficiency and accuracy of due diligence material review can be improved, the need for manual intervention can be reduced, and the review process can be ensured to cover all necessary checkpoints, effectively preventing risks.

[0064] Optionally, in the enterprise investigation material quality inspection method provided in this application embodiment, polling and detecting the task execution status and quality inspection results of each quality inspection task to obtain the target detection result includes: polling and detecting whether the task execution status of each quality inspection task is completed to obtain an initial detection result; if the initial detection result indicates that the task execution status of all quality inspection tasks is completed, obtaining the quality inspection results generated by all quality inspection tasks to obtain the target detection result; if the initial detection result indicates that the task execution status of any quality inspection task is in progress, obtaining the task execution time of that quality inspection task; if the task execution time of that quality inspection task exceeds a preset time threshold, terminating the quality inspection task and generating an anomaly diagnosis log for that quality inspection task; and obtaining the target detection result based on the anomaly diagnosis log and all generated quality inspection results.

[0065] In this embodiment, periodic status checks can be performed on each quality inspection task to monitor its progress in real time and confirm whether the task has been completed. In an optional example, this can be done by querying the background task management status database to obtain the current status of each task (running, completed, failed, etc.).

[0066] By polling, an initial inspection result can be obtained, which can be used to indicate the execution status of all quality inspection tasks. If the status of all tasks is "completed", the results of each quality inspection task can be collected. These results can include specific information such as whether the quality inspection item corresponding to the quality inspection task passed, whether a problem was found, and the risk assessment level. The collection of all quality inspection results constitutes the target inspection result, providing a basis for subsequent business decisions.

[0067] If any quality inspection task is found to be still in the "running" state during the polling process, its runtime can be further checked. If the execution time of a quality inspection task exceeds a preset time threshold (e.g., 10 minutes), it can be determined that the quality inspection task is in a processing bottleneck or abnormal state. The quality inspection task can be automatically terminated, and an anomaly diagnosis log can be generated. The anomaly diagnosis log can include the reason for the task termination and related information to facilitate subsequent troubleshooting and system optimization.

[0068] Even if some quality inspection tasks are terminated due to timeout after the quality inspection tasks are completed, in this embodiment, the final target detection result can be generated based on the existing quality inspection results and anomaly diagnosis logs to reflect the completion status of the entire quality inspection process, as well as any problems or anomalies encountered in the process, providing business personnel with comprehensive risk assessment and decision support.

[0069] This embodiment enables the automated execution of multiple quality inspection tasks and intelligent handling of abnormal situations during task execution, ensuring the efficiency, accuracy, and completeness of the quality inspection process. It also provides necessary data support for subsequent process optimization and system upgrades.

[0070] Optionally, in the enterprise investigation material quality inspection method provided in this application embodiment, the due diligence materials include: textual materials and image materials. The image materials include: videos representing the business intentions of the target enterprise and on-site images of the target enterprise. Each quality inspection task is executed, including: determining the quality inspection service to be invoked for each quality inspection task. The types of quality inspection services include: information quality inspection service, video quality inspection service, and image quality inspection service. The information quality inspection service is used to detect whether the textual materials are authentic using a large language model. The video quality inspection service is used to detect whether the business intentions expressed in the video are consistent with the business application using a large language model. The image quality inspection service is used to detect whether the business premises of the target enterprise are authentic using a large language model. The quality inspection service associated with each quality inspection task is invoked, and the due diligence materials are inspected using a large language model.

[0071] The aforementioned information quality inspection service utilizes a large language model to detect the authenticity of textual materials. This service can include address information inspection and contact information inspection. The address information inspection service automatically cleans up discrepancies in address text using a large language model and verifies the consistency between the due diligence address (i.e., the target company's address in the due diligence materials) and the company's registered address (i.e., the target company's actual registered address). The contact information inspection service automatically verifies whether contact information matches according to business rules.

[0072] The aforementioned video quality inspection service can be used to detect whether the business intentions expressed in the video are consistent with the business application using a large language model. For example, the video quality inspection service can extract the video audio content in real time through deep fusion speech recognition, combine it with a large language model for semantic analysis and intent recognition, and automatically judge the consistency between the video statement and the business script text to ensure the authenticity of the content.

[0073] The aforementioned image quality inspection service can be used to detect whether the business premises of a target enterprise are genuine using large language models. For example, by adopting a differentiated processing strategy, standard certificates such as ID cards and business licenses can be identified using OCR (Optical Character Recognition) to extract image information into structured data for intelligent verification. For on-site images of enterprises, multimodal large models can be used to accurately identify visual elements such as enterprise logos and doorplate information, and automatically cross-verify them with data such as office addresses and enterprise names submitted by customers to effectively identify the risk of fake business premises.

[0074] When performing quality inspection tasks, the corresponding quality inspection service can be invoked according to the nature and requirements of each quality inspection item. For example, when the quality inspection task involves "address information consistency", the information quality inspection service can be invoked, using a large language model to process address text; if the quality inspection task involves "business intention video authenticity check", the video quality inspection service can be invoked, using speech recognition and semantic analysis to determine the consistency of video content; when the quality inspection task involves "compliance of enterprise on-site images", the image quality inspection service can be invoked, using image recognition technology to detect key visual elements in the image.

[0075] By conducting in-depth analysis of various types of due diligence materials, we not only improved the efficiency of the review process but also enhanced our ability to judge the authenticity of the materials, effectively preventing potential risks of illegal transactions and ensuring that the business processes of financial institutions are more secure, compliant, and efficient.

[0076] Optionally, in the enterprise investigation material quality inspection method provided in the embodiments of this application, after determining N quality inspection tasks based on N quality inspection items and due diligence materials, the method further includes: adding the N quality inspection tasks to a target message queue; and executing each quality inspection task, including: reading each quality inspection task from the target message queue and executing the read quality inspection task.

[0077] The aforementioned target message queue can serve as a task scheduler, capable of storing and managing pending quality inspection tasks in a certain order or priority.

[0078] After determining the quality inspection tasks to be performed based on N quality inspection items and due diligence materials, these tasks can be encapsulated into messages and added to the target message queue. The target message queue can also act as a decoupling tool between different components, eliminating the need for direct communication between the producers of quality inspection tasks (e.g., the front end of enterprise investigation material quality inspection) and the consumers (e.g., the service that performs the quality inspection tasks), thereby reducing system complexity and improving stability and scalability.

[0079] In this embodiment, the message queue can be continuously polled to find unprocessed quality inspection tasks. Once a quality inspection task is read, it can be retrieved from the target message queue and the corresponding quality inspection service can be executed. For example, OCR recognition technology can be used to parse text information in an image, natural language processing technology can be used to analyze audio content in a video, or the consistency and compliance of information can be verified.

[0080] Managing task flows based on target message queues ensures smooth task execution even under high network concurrency or limited system resources. Furthermore, placing quality inspection tasks into message queues facilitates retry mechanisms and fault isolation in case of task execution anomalies. This also allows for the addition of more quality inspection services or adjustments to task processing logic to meet evolving business needs.

[0081] Optionally, in the enterprise investigation material quality inspection method provided in this application embodiment, after performing quality inspection on the due diligence materials based on N quality inspection items using a large language model to obtain the target quality inspection result, the method further includes: based on the target quality inspection result, marking the quality inspection result of each quality inspection item to obtain N marked quality inspection items; visually displaying the N marked quality inspection items, and receiving feedback information from the target object on the N marked quality inspection items, wherein the target object includes: the object in the financial institution that processes business applications for the target enterprise, and the feedback information is used to indicate whether the quality inspection result of each quality inspection item is accurate.

[0082] The aforementioned target quality inspection results may include: the quality inspection results associated with each quality inspection item. The quality inspection results associated with each quality inspection item may include, but are not limited to: pass, fail, quality inspection task execution failure, and further review required.

[0083] In this embodiment, the quality inspection results associated with each quality inspection item can be marked for that quality inspection item.

[0084] The quality inspection items, after being marked with N tags, can be displayed visually to the target audience (e.g., business processing personnel in financial institutions) through a page on an electronic screen. For example, they can be presented in the form of a list, chart, or other intuitive interface, allowing the target audience to see the status of each quality inspection item at a glance.

[0085] After viewing the visualized quality inspection results, the target audience (e.g., business processing personnel) can provide feedback on the results of each quality inspection item. This feedback is used to assess the accuracy of the large language model's judgments; for example, users can evaluate the effectiveness of the quality inspection results by giving likes or dislikes.

[0086] It should be noted that the collected feedback information can be used to adjust the parameters or rules of the large language model, so that the model can make more accurate judgments when processing similar materials.

[0087] In this embodiment, providing feedback on the quality inspection results of the quality inspection items can ensure that the due diligence material quality inspection process is more aligned with actual business needs, thereby improving the overall review quality and customer satisfaction.

[0088] Optionally, in the enterprise investigation material quality inspection method provided in this application embodiment, N quality inspection items are determined based on the business application, including: generating a quality inspection list based on the business application, wherein the quality inspection list includes: M quality inspection items, where M is an integer greater than N; displaying the quality inspection list on a target interface, and receiving the N quality inspection items selected by the target object using the target interface, wherein the target interface includes at least: M quality inspection items, and check buttons for the M quality inspection items.

[0089] Upon receiving a business application from a target company, a quality control checklist can be generated. This checklist may include all potentially applicable quality control items. These M checklist items may include, but are not limited to: verifying address information, validating contact information, verifying the authenticity of financial statements, reviewing the content of videos expressing business intent, and checking the compliance of image files. Each checklist item is designed to ensure that a specific aspect of the application materials meets the financial institution's business compliance requirements.

[0090] The generated quality inspection list can be displayed on a target interface, which can be an operating platform designed for the target object (e.g., business processing personnel) to receive instructions and feedback from the target object. The target interface can list all M quality inspection items in the quality inspection list, and each quality inspection item can have a checkbox next to it, so that the target object can easily browse all possible quality inspection items and select the specific quality inspection items to be performed, thus determining N quality inspection items.

[0091] Once N quality inspection items have been selected for the target object and the selected quality inspection items have been submitted, the corresponding quality inspection tasks can be generated and executed.

[0092] By providing customized review solutions for each business application through quality inspection categories, and by involving the target audience, we ensure that the review focus is highly aligned with actual business needs, thereby improving the efficiency and accuracy of the entire review process.

[0093] In this embodiment, based on technologies such as OCR recognition and large-scale artificial intelligence models, the collaborative analysis of multimodal data (text, video, images, and other due diligence materials) transforms the highly manual and time-consuming review process in the due diligence workflow into AI-assisted quality inspection. This significantly reduces review time and effectively improves the overall quality and efficiency of due diligence, saving labor costs while ensuring review effectiveness. Intelligent quality inspection also greatly reduces subjective errors and omissions that are unavoidable in manual verification, significantly improving the identification rate and coverage of potential risks, contradictory information, and compliance vulnerabilities. This effectively enhances the accuracy of risk identification and prevention capabilities, thereby improving the accuracy and reliability of due diligence results.

[0094] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0095] Example 2

[0096] Embodiment 2 of this application provides an optional enterprise survey material quality inspection system, which can be used to execute the enterprise survey material quality inspection method provided in Embodiment 1 of this invention.

[0097] The enterprise investigation material quality inspection system provided in Embodiment 2 of this application can utilize large-scale model technologies such as intelligent video analysis and multimodal content review of text and images to achieve intelligent inspection of due diligence. Through a combination of pre-inspection, real-time monitoring during the process, and regular post-inspection, after the customer service (e.g., the target enterprise) uploads due diligence materials such as videos verifying business intentions, the target entity (e.g., the account manager) can use the enterprise investigation material quality inspection system to verify the authenticity and accuracy of customer information and monitor the clarity and compliance of image files.

[0098] Figure 3 This is a schematic diagram of an enterprise investigation material quality inspection system provided according to an embodiment of this application, such as... Figure 3 As shown, it includes: a data pre-storage model, a quality inspection service module, a task execution module, a result polling model, and a result feedback module.

[0099] (a) Data Pre-storage Module:

[0100] After receiving a business application from a customer (e.g., a target company), a financial institution branch can enter customer information and upload due diligence materials into the system. These materials include videos of the customer's business intentions and images taken during on-site due diligence (e.g., photos taken by financial institution personnel at the target company's office). The system supports the unified storage of both structured and unstructured data. The data pre-storage module can accurately link discrete image data with customer information, while simultaneously implementing automated document integrity checks and basic compliance screenings to ensure the validity and traceability of uploaded materials, providing a data foundation for quality inspection tasks.

[0101] (II) Quality Inspection Service Module:

[0102] The quality inspection service module can adopt an atomic service architecture to achieve functional decoupling and flexible assembly, transforming manual verification actions in related technologies into automated quality inspection items. Each quality inspection item can correspond to an independent microservice unit. Business personnel can flexibly select one or more quality inspection items in a visual configuration interface according to different due diligence needs, meeting diverse quality inspection requirements.

[0103] The quality inspection service module can include various quality inspection functions such as information quality inspection, video quality inspection, and image quality inspection. Among them, the address information quality inspection service can automatically clean the expression differences in the address text through a large model (i.e., a large language model) and verify the consistency between the due diligence address and the company's registered address; the contact information quality inspection service automatically verifies whether the contact information matches according to business rules.

[0104] Video quality inspection services can extract video audio content in real time through deep integration of speech recognition, combine it with large models for semantic analysis and intent recognition, and automatically judge the consistency between video statements and business script text to ensure the authenticity of the content.

[0105] Image quality inspection services can employ differentiated processing strategies. For standard certificates such as ID cards and business licenses, OCR recognition is used to extract structured data from the images for intelligent verification. For on-site images of businesses, multimodal large-scale models are used to accurately identify visual elements such as company logos and doorplate information, and automatic cross-validation is performed with data such as the office address and company name submitted by the client to effectively identify the risk of fraudulent business locations.

[0106] (III) Task Execution Module:

[0107] Financial institution personnel can select a combination of quality inspection items based on the due diligence scenario requirements. The quality inspection execution module can then automatically trigger service orchestration, dynamically retrieving the corresponding quality inspection configuration parameters and pre-stored due diligence data according to preset rules, and triggering the execution of the quality inspection task. For computationally intensive tasks such as video analysis and image recognition, an asynchronous processing mechanism is enabled, with execution results obtained through result callbacks.

[0108] (iv) Result Polling Module:

[0109] The result polling module can build a closed-loop monitoring system for the entire process, continuously polling the execution progress of each quality inspection item and updating its status information. If a quality inspection item executes successfully, the execution result is recorded. When all quality inspection items return results, a quality inspection conclusion can be automatically generated and the task completion status can be marked. To avoid process blockages caused by system anomalies, the result polling module can be configured with a timeout circuit breaker mechanism. If any quality inspection item in a single quality inspection task fails to return a result for 10 consecutive minutes, the system can automatically terminate the task flow, mark it as a timeout failure, and generate an anomaly diagnostic log, preventing long-term resource occupation and ensuring business closure.

[0110] (v) Results Feedback Module:

[0111] After quality inspection is completed, business personnel can evaluate the effectiveness of the inspection results through a visual interface by liking or disliking them. The results feedback module triggers a quality inspection rule review process for frequently disputed items and automatically incorporates the corrected samples into the model retraining dataset, driving model upgrades and iterations.

[0112] Figure 4 This is a flowchart of the quality inspection process for enterprise investigation materials provided in the embodiments of this application, such as... Figure 4 As shown, it includes:

[0113] Step 401: The user (e.g., a business person from a financial institution) enters key business information and basic customer data on the designated page of the system based on the materials submitted by the customer.

[0114] Step 402: The user uploads and adds the customer's image materials, including multi-angle photos taken on site and videos of the customer's business intentions recorded by the customer, to the corresponding customer due diligence task.

[0115] Step 403: Based on the due diligence scenario requirements and risk concerns, users can flexibly select one or more intelligent quality inspection items from the list of quality inspection items provided on the page.

[0116] Step 404: After the user confirms the information, click to submit the quality inspection task. The system background will automatically trigger task orchestration, obtain the configuration parameters and due diligence data of the corresponding quality inspection items, and trigger task execution.

[0117] Step 405: Based on the tasks triggered in the background, the system calls the corresponding quality inspection services to execute the task processing flow. By continuously polling the execution progress of each quality inspection item, the system updates the status of completed tasks in real time and performs circuit breaking for timed-out tasks, ultimately achieving a closed-loop process.

[0118] Step 406: After the intelligent quality inspection task is completed, users can view the details of each quality inspection result on the page, including whether the quality inspection passed, abnormal information, risk warnings, etc.

[0119] Step 407: After carefully reviewing the quality inspection results, users can quickly and intuitively evaluate the effectiveness of the intelligent quality inspection results provided by the system by "liking" (for accurate results) or "disliking" (for incorrect results).

[0120] Step 408: After confirming the results, the user clicks the submit button to formally submit the due diligence task, which includes the original materials, entered information, and quality inspection results, to the next processing stage.

[0121] Through the above steps, the manual review process in related technologies can be upgraded to AI-assisted quality inspection. After users upload the materials to be reviewed and select the required due diligence items, the system can intelligently review the due diligence risks. This saves labor costs, significantly improves the accuracy of risk identification, and increases business processing efficiency. In addition, the system can continuously drive model optimization based on user feedback data, achieving an intelligent upgrade of the due diligence process.

[0122] Example 3

[0123] This application also provides a quality inspection device for enterprise survey materials. It should be noted that this device can be used to execute the quality inspection method for enterprise survey materials provided in this application. The following describes the quality inspection device for enterprise survey materials provided in this application.

[0124] According to an embodiment of this application, an apparatus for implementing the above-described method for quality inspection of enterprise survey materials is also provided, such as... Figure 5 As shown, the device includes: an acquisition unit 51, a first determination unit 52, a quality inspection unit 53, and a second determination unit 54.

[0125] Among them, the acquisition unit 51 is used to acquire the due diligence materials of the target company when the target company submits a business application to the financial institution.

[0126] The first determining unit 52 is used to determine N quality inspection items based on the business application, where N is a positive integer;

[0127] Quality inspection unit 53 is used to perform quality inspection on due diligence materials based on N quality inspection items and using a large language model to obtain the target quality inspection results.

[0128] The second determining unit 54 is used to determine the processing method of the business application based on the target quality inspection results.

[0129] In the enterprise investigation material quality inspection device provided in this application embodiment, the acquisition unit 51 can acquire the due diligence materials of the target enterprise when the target enterprise submits a business application to a financial institution. The first determining unit 52 determines N quality inspection items based on the business application, where N is a positive integer. The quality inspection unit 53 uses a large language model to inspect the due diligence materials based on the N quality inspection items, obtaining the target quality inspection result. The second determining unit 54 determines the processing method for the business application based on the target quality inspection result. This solves the technical problem of low efficiency in manual quality inspection of due diligence materials during the processing of enterprise business in related technologies. In this embodiment, by setting multiple quality inspection items and using a large language model to inspect the due diligence materials, the low efficiency and high error rate of manual quality inspection of due diligence materials in related technologies are avoided, thereby achieving the technical effect of improving the efficiency and accuracy of due diligence material quality inspection.

[0130] Optionally, in the enterprise investigation material quality inspection device provided in this application embodiment, the quality inspection unit includes: a first determining subunit, used to determine N quality inspection tasks based on N quality inspection items and due diligence materials, wherein each quality inspection task is used to perform quality inspection on the materials to be inspected by the quality inspection task using a large language model, and each quality inspection task includes: configuration parameters associated with one of the quality inspection items and the materials to be inspected by the quality inspection item; a first processing subunit, used to execute each quality inspection task and poll and detect the task execution status and quality inspection results of each quality inspection task to obtain a target detection result; and a second determining subunit, used to determine a target quality inspection result based on the target detection result.

[0131] Optionally, in the enterprise investigation material quality inspection device provided in this application embodiment, the first processing subunit includes: a detection module, used to poll and detect whether the task execution status of each quality inspection task is completed, and obtain an initial detection result; an acquisition module, used to acquire the quality inspection results generated by all quality inspection tasks when the initial detection result indicates that the task execution status of all quality inspection tasks is completed, and obtain a target detection result; and a first processing module, used to acquire the task execution time of any quality inspection task when the initial detection result indicates that the task execution status of any quality inspection task is in progress, terminate the quality inspection task when the task execution time of the quality inspection task exceeds a preset time threshold, generate an abnormal diagnosis log for the quality inspection task, and obtain the target detection result based on the abnormal diagnosis log and all generated quality inspection results.

[0132] Optionally, in the enterprise investigation material quality inspection device provided in this application embodiment, the due diligence materials include: textual materials and image data. The image data includes: videos representing the business intentions of the target enterprise and on-site images of the target enterprise. The processing subunit further includes: a determination module, used to determine the quality inspection service to be invoked for each quality inspection task. The types of quality inspection services include: information quality inspection service, video quality inspection service, and image quality inspection service. The information quality inspection service is used to detect whether the textual materials are authentic using a large language model. The video quality inspection service is used to detect whether the business intentions expressed in the video are consistent with the business application using a large language model. The image quality inspection service is used to detect whether the business premises of the target enterprise are authentic using a large language model. The second processing module is used to invoke the quality inspection service associated with each quality inspection task and to perform quality inspection on the due diligence materials using a large language model.

[0133] Optionally, in the enterprise investigation material quality inspection device provided in the embodiments of this application, the enterprise investigation material quality inspection device includes: an adding unit, used to add N quality inspection tasks to a target message queue after determining N quality inspection tasks based on N quality inspection items and due diligence materials; and a processing subunit including: an execution module, used to read each quality inspection task from the target message queue and execute the read quality inspection task.

[0134] Optionally, in the enterprise investigation material quality inspection device provided in this application embodiment, the enterprise investigation material quality inspection device further includes: a marking unit, used to mark the quality inspection result of each quality inspection item based on the target quality inspection result after the due diligence materials are inspected using a large language model based on N quality inspection items, and N marked quality inspection items are obtained; and a processing unit, used to visualize the N marked quality inspection items and receive feedback information from the target object on the N marked quality inspection items, wherein the target object includes: the object in the financial institution that processes business applications for the target enterprise, and the feedback information is used to indicate whether the quality inspection result of each quality inspection item is accurate.

[0135] Optionally, in the enterprise investigation material quality inspection device provided in this application embodiment, the first determining unit includes: a generation subunit, used to generate a quality inspection list based on a business application, wherein the quality inspection list includes: M quality inspection items, where M is an integer greater than N; and a second processing subunit, used to display the quality inspection list on a target interface, and to receive N quality inspection items selected by the target object using the target interface, wherein the target interface includes at least: M quality inspection items, and check buttons for the M quality inspection items.

[0136] It should be noted that the acquisition unit 51, the first determination unit 52, the quality inspection unit 53, and the second determination unit 54 mentioned above correspond to steps S201 to S204 in Embodiment 1. Each unit and the corresponding step implement the same instance and application scenario, but are not limited to the content disclosed in Embodiment 1. It should be noted that the above modules or units can be hardware components or software components stored in memory (e.g., memory 104) and processed by one or more processors (e.g., processors 102a, 102b, ..., 102n). The above modules can also be part of a device and run in the computer terminal 10 provided in Embodiment 1.

[0137] Example 4

[0138] Embodiments of this application may provide an electronic device. Figure 6 This is a structural block diagram of an electronic device according to an embodiment of this application. Figure 6 As shown, the electronic device may include: one or more ( Figure 6 (Only one is shown) Processor 602, memory 604, memory controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0139] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the methods and apparatus in the embodiments of this application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby implementing the above-described methods. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0140] The processor can access information and applications stored in memory via a transmission device to perform the following steps: when the target company submits a business application to a financial institution, obtain the target company's due diligence materials; based on the business application, determine N quality inspection items, where N is a positive integer; based on the N quality inspection items, use a large language model to perform quality inspection on the due diligence materials to obtain the target quality inspection result; based on the target quality inspection result, determine the processing method for the business application.

[0141] The processor can also access information and applications stored in memory via a transmission device to perform the following steps: Based on N quality inspection items, perform quality inspection on due diligence materials using a large language model to obtain a target quality inspection result, including: Based on the N quality inspection items and due diligence materials, determine N quality inspection tasks, where each quality inspection task is used to perform quality inspection on the materials inspected by the large language model, and each quality inspection task includes: configuration parameters associated with one of the quality inspection items and the materials inspected by that quality inspection item; execute each quality inspection task and poll the task execution status and quality inspection results of each quality inspection task to obtain a target detection result; determine the target quality inspection result based on the target detection result.

[0142] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: polling and detecting the task execution status and quality inspection results of each quality inspection task to obtain the target detection result, including: polling and detecting whether the task execution status of each quality inspection task is completed to obtain the initial detection result; if the initial detection result indicates that the task execution status of all quality inspection tasks is completed, obtaining the quality inspection results generated by all quality inspection tasks to obtain the target detection result; if the initial detection result indicates that the task execution status of any quality inspection task is in progress, obtaining the task execution time of that quality inspection task; if the task execution time of that quality inspection task exceeds a preset time threshold, terminating the quality inspection task and generating an anomaly diagnosis log for that quality inspection task; and obtaining the target detection result based on the anomaly diagnosis log and all generated quality inspection results.

[0143] The processor can also invoke information and applications stored in the memory via a transmission device to perform the following steps: The due diligence materials include: textual materials and image materials. The image materials include: videos representing the target company's business intentions and on-site images of the target company. Each quality inspection task is executed, including: determining the quality inspection services to be invoked for each quality inspection task. The types of quality inspection services include: information quality inspection services, video quality inspection services, and image quality inspection services. The information quality inspection service is used to detect the authenticity of the textual materials using a large language model. The video quality inspection service is used to detect whether the business intentions expressed in the video are consistent with the business application using a large language model. The image quality inspection service is used to detect whether the target company's business premises are real using a large language model. The quality inspection service associated with each quality inspection task is invoked, and the due diligence materials are inspected using a large language model.

[0144] The processor can also invoke information and applications stored in the memory via the transmission device to perform the following steps: after determining N quality inspection tasks based on N quality inspection items and due diligence materials, the process further includes: adding the N quality inspection tasks to the target message queue; and executing each quality inspection task, including: reading each quality inspection task from the target message queue and executing the read quality inspection task.

[0145] The processor can also access information and applications stored in the memory via a transmission device to perform the following steps: After performing quality inspection on due diligence materials based on N quality inspection items using a large language model and obtaining the target quality inspection result, the processor further includes: based on the target quality inspection result, marking the quality inspection result of each quality inspection item to obtain N marked quality inspection items; visually displaying the N marked quality inspection items and receiving feedback information from the target object on the N marked quality inspection items, wherein the target object includes: the object in the financial institution that processes business applications for the target enterprise, and the feedback information is used to indicate whether the quality inspection result of each quality inspection item is accurate.

[0146] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: Based on the business application, determine N quality inspection items, including: Based on the business application, generate a quality inspection list, wherein the quality inspection list includes: M quality inspection items, where M is an integer greater than N; Display the quality inspection list on the target interface, and use the target interface to receive the N quality inspection items selected by the target object, wherein the target interface includes at least: M quality inspection items, and check buttons for the M quality inspection items.

[0147] By adopting the embodiments of this application, multiple quality inspection items are set, and a large language model is used to inspect the due diligence materials. This avoids the situation in related technologies where manual inspection of due diligence materials is inefficient and has a high error rate, thereby achieving the technical effect of improving the quality inspection efficiency and accuracy of due diligence materials.

[0148] Those skilled in the art will understand that Figure 6 The structure shown is for illustrative purposes only. Electronic devices can also be smartphones, tablets, handheld computers, mobile internet devices (MIDs), PADs, and other terminal devices. Figure 6 This does not limit the structure of the aforementioned electronic device. For example, electronic devices may also include components that are more... Figure 6 The more or fewer components shown (such as network interfaces, display devices, etc.), or having the same Figure 6 The different configurations shown.

[0149] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0150] Example 5

[0151] Embodiments of this application also provide a storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the enterprise investigation material quality inspection method provided in Embodiment 1.

[0152] Optionally, in this embodiment, the storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any mobile terminal in a group of mobile terminals.

[0153] This application also provides a computer program product that, when executed on a data processing device, is suitable for performing steps of a method for quality inspection of enterprise survey materials.

[0154] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0155] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0156] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0157] The units described as separate components may or may not be physically separate. The components shown as units 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 units can be selected to achieve the purpose of this embodiment according to actual needs.

[0158] Furthermore, the functional units in the various embodiments of this application 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 as a software functional unit.

[0159] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0160] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for quality inspection of enterprise survey materials, characterized in that, include: When the target company submits a business application to a financial institution, obtain the due diligence materials of the target company; Based on the aforementioned business application, N quality inspection items are determined, where N is a positive integer; Based on N quality inspection items, a large language model is used to conduct quality inspection on the due diligence materials to obtain the target quality inspection results. Based on the target quality inspection results, the processing method for the business application is determined.

2. The quality inspection method according to claim 1, characterized in that, Based on N quality inspection items, a large language model is used to perform quality inspection on the due diligence materials to obtain the target quality inspection results, including: Based on the N quality inspection items and the due diligence materials, N quality inspection tasks are determined. Each quality inspection task is used to inspect the materials to be inspected by the large language model. Each quality inspection task includes: configuration parameters associated with one of the quality inspection items and the materials to be inspected by the quality inspection item. Execute each of the quality inspection tasks, and poll the task execution status and quality inspection results of each of the quality inspection tasks to obtain the target inspection result; Based on the target detection results, the target quality inspection results are determined.

3. The quality inspection method according to claim 2, characterized in that, Polling the execution status and results of each quality inspection task yields the target inspection results, including: Poll the execution status of each quality inspection task to check whether it has been completed, and obtain the initial detection result; If the initial detection result indicates that the task execution status of all the quality inspection tasks is completed, the quality inspection results generated by all the quality inspection tasks are obtained to obtain the target detection result; If the initial detection result indicates that the task execution status of any of the quality inspection tasks is in progress, the task execution time of the quality inspection task is obtained. If the task execution time of the quality inspection task exceeds a preset time threshold, the quality inspection task is terminated, and an anomaly diagnosis log of the quality inspection task is generated. Based on the anomaly diagnosis log and all generated quality inspection results, the target detection result is obtained.

4. The quality inspection method according to claim 2, characterized in that, The due diligence materials include: written materials and video materials. The video materials include: videos representing the target company's business intentions and on-site images of the target company. Performing each quality inspection task includes: The quality inspection services to be invoked for each quality inspection task are determined. The types of quality inspection services include: information quality inspection service, video quality inspection service, and image quality inspection service. The information quality inspection service is used to detect whether the textual materials are authentic using the large language model. The video quality inspection service is used to detect whether the business intention expressed in the video is consistent with the business application using the large language model. The image quality inspection service is used to detect whether the business premises of the target enterprise are authentic using the large language model. Each quality inspection task invokes the associated quality inspection service and uses the large language model to perform quality inspection on the due diligence materials.

5. The quality inspection method according to claim 2, characterized in that, The method further includes: After determining N quality inspection tasks based on the N quality inspection items and the due diligence materials, the method further includes: adding the N quality inspection tasks to the target message queue. Executing each of the quality inspection tasks includes: reading each of the quality inspection tasks from the target message queue and executing the read quality inspection tasks.

6. The quality inspection method according to claim 1, characterized in that, After performing quality checks on the due diligence materials based on N quality inspection items using a large language model to obtain the target quality inspection results, the process also includes: Based on the target quality inspection results, the quality inspection results of each quality inspection item are marked, resulting in N marked quality inspection items; The system visualizes the N marked quality inspection items and receives feedback from the target object regarding the N marked quality inspection items. The target object includes the financial institution that processes the business application for the target enterprise. The feedback information is used to indicate whether the quality inspection result of each quality inspection item is accurate.

7. The quality inspection method according to claim 6, characterized in that, Based on the aforementioned business application, N quality inspection items are determined, including: Based on the business application, a quality inspection list is generated, wherein the quality inspection list includes M quality inspection items, where M is an integer greater than N; The quality inspection list is displayed on the target interface, and the target interface is used to receive N quality inspection items selected by the target object. The target interface includes at least M quality inspection items and M check buttons for the quality inspection items.

8. A quality inspection device for enterprise survey materials, characterized in that, include: The acquisition unit is used to acquire the due diligence materials of the target company when the target company submits a business application to a financial institution. The first determining unit is used to determine N quality inspection items based on the business application, where N is a positive integer; The quality inspection unit is used to perform quality inspection on the due diligence materials based on N quality inspection items and using a large language model to obtain the target quality inspection result. The second determining unit is used to determine the processing method of the business application based on the target quality inspection results.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored executable program, wherein, when the executable program is executed, it controls the device on which the computer-readable storage medium is located to perform the enterprise investigation material quality inspection method according to any one of claims 1 to 7.

10. A computer program product comprising computer instructions, characterized in that, When the computer instructions are executed by the processor, they implement the steps of the enterprise investigation material quality inspection method according to any one of claims 1 to 7.

Images