Test method, system, device and medium for picture-text data auditing function
By constructing a multi-dimensional test case matrix and an automated review process, the problem of insufficient coverage in the testing of text and image data review functions of ride-hailing platforms was solved, achieving accurate verification and efficient testing throughout the entire process.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- BEIJING BAIJU YIXING TECH CO LTD
- Filing Date
- 2026-01-21
- Publication Date
- 2026-05-29
AI Technical Summary
In existing technologies, the testing of image and text data verification functions on ride-hailing platforms suffers from insufficient test coverage and poor comprehensiveness of verification results, making it difficult to meet complex testing requirements.
Construct a test case matrix that includes dimensions such as document type, data quality, and information matching degree. Generate text data and simulated document image data. Simulate the review process through automated scripts, obtain and verify the review flow data, and achieve coverage of multi-dimensional test scenarios and full-process verification.
It improves the comprehensiveness and accuracy of test coverage, enables the discovery of anomalies in the review process, reduces the cost of manually generating test data, and improves testing efficiency and the comprehensiveness of verification.
Smart Images
Figure CN122111843A_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to the field of testing technology, and in particular to a testing method, system, device and medium for a text and image data review function. Background Technology
[0002] With the rapid development of the internet-based transportation industry, ride-hailing platforms and other service providers are increasingly stringent in their qualification verification of drivers. Among these verifications, the review of textual and image data (such as vehicle registration certificates and driver's licenses) is a core element in ensuring the platform's operational compliance and service security. The stability and accuracy of the textual and image data verification function are directly related to the platform's operational risk control; therefore, testing and verifying this function is crucial.
[0003] In related technologies, testing of the image and text data verification function of ride-hailing platforms often adopts traditional interface automation testing or UI-based automated testing solutions. However, these traditional testing solutions have many shortcomings and are difficult to meet the complex testing requirements of image and text data verification functions, resulting in insufficient test coverage and poor comprehensiveness of verification results. Summary of the Invention
[0004] Therefore, it is necessary to provide a testing method, system, device, and medium that can improve the comprehensiveness of the testing of graphic and text data review functions, addressing at least one of the aforementioned technical problems.
[0005] In a first aspect, embodiments of this disclosure provide a testing method for a text and image data review function, the method including the following steps: Based on the image and text data verification business logic of the ride-hailing platform under test, a test case matrix is constructed. According to the construction dimension information of each test case in the test case matrix, test data including text data and simulated document image data is generated. The text data and simulated document image data are associated to form a test work order, which is sent to the ride-hailing platform under test. The verification flow data of the ride-hailing platform under test for the test work order is obtained. The verification flow data is verified according to the expected verification results to obtain the test results of the image and text data verification function of the ride-hailing platform under test.
[0006] The test case matrix is constructed in at least the dimensions of document type, data quality, and information matching degree. The test case matrix also includes the expected review result for each test case. Data quality is used to represent the quality of the text and image data to be reviewed corresponding to the test case, and information matching degree is used to represent the degree of matching between the text and image data to be reviewed and the benchmark verification data.
[0007] The audit process data includes audit status data, log data, and message queue data.
[0008] In some embodiments, document types include ID cards, vehicle registration certificates, and driver's licenses. Data quality includes a first quality dimension representing the identifiability of the text and image data to be reviewed, a second quality dimension representing authenticity, and a third quality dimension representing timeliness. Information matching degree includes exact match, partial match, and no match.
[0009] In some embodiments, generating test data including text data and simulated document image data based on the construction dimension information of each test case in the test case matrix may include the following steps: calling a preset text data generation tool to generate structured text data according to the document type and information matching degree corresponding to the test case; and calling a preset image data generation tool to generate simulated document image data according to the document type and data quality corresponding to the test case.
[0010] In some embodiments, after sending the test data to the platform under test, the method may further include the following steps: The system uses a pre-set automated script or a pre-set application programming interface to simulate an auditor logging into the ride-hailing platform under test, triggering the platform's image and text data auditing function to audit the test data.
[0011] In some embodiments, the test results include state verification results, log verification results, and message queue verification results.
[0012] Verifying the review process data based on the expected review results, and obtaining the test results of the image and text data review function of the ride-hailing platform under test, may include the following steps: Verify whether the audit status data matches the expected audit flow status in the expected audit results to obtain the status verification result; verify whether the log record data corresponds to the expected audit results to obtain the log verification result; verify whether the message queue data corresponds to the expected audit results to obtain the message queue verification result.
[0013] In some embodiments, the method may further include: generating a test report based on the test results corresponding to each test case in the test case matrix, wherein the test report is used to record the test case pass rate, defect log and / or test work order status flow information in the test results.
[0014] In some embodiments, the method may further include the following steps: in response to an audit business rule change instruction, obtaining the target business rule corresponding to the change instruction, and updating the test case matrix according to the target business rule.
[0015] In a second aspect, embodiments of this disclosure provide a testing system for graphic data review functions. The system includes a test management platform, a test data generation engine, an automated execution engine, a process status monitor, and a verification center.
[0016] The test management platform is used to review the business logic based on the graphic and textual data of the ride-hailing platform under test and to build a test case matrix.
[0017] The test case matrix is constructed in at least the dimensions of document type, data quality, and information matching degree. The test case matrix also includes the expected review result for each test case. Data quality is used to indicate the clarity of the text and image data to be reviewed corresponding to the test case, and information matching degree is used to indicate the degree of matching between the text and image data to be reviewed and the benchmark verification data.
[0018] The test data generation engine is used to generate test data, including text data and simulated document image data, based on the construction dimension information of each test case in the test case matrix.
[0019] An automated execution engine is used to associate text data and simulated document image data to form a test work order and send it to the ride-hailing platform under test.
[0020] The process status monitor is used to obtain the review and processing data of the test work orders from the ride-hailing platform under test.
[0021] The audit process data includes audit status data, log data, and message queue data.
[0022] The verification center is used to verify the review process data based on the expected review results, and to obtain the test results of the image and text data review function of the ride-hailing platform under test.
[0023] In a third aspect, embodiments of this disclosure provide a computer device including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of a test method for the image and text data review function provided in any embodiment of the first aspect of this disclosure.
[0024] In a fourth aspect, embodiments of this disclosure provide a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the test method for the graphic data review function provided in any embodiment of the first aspect of this disclosure.
[0025] The aforementioned testing methods, systems, equipment, and media for the text and image data verification function achieve a systematic review and coverage of multi-dimensional testing scenarios by constructing a test case matrix that includes information on document type, data quality, and information matching degree. Compared to the fragmented test case design in related technologies, this method can comprehensively cover multi-dimensional combinations of scenarios with different document types, data qualities, and information matching degrees, improving the comprehensiveness of test coverage and effectively avoiding test vulnerabilities caused by scenario omissions. Based on the dimensional information of the test case construction, test data containing text data and simulated document image data is automatically generated. This can accurately generate simulated document image data and corresponding text data that meet the requirements of different testing scenarios, ensuring the authenticity and diversity of test data. It can better simulate the data characteristics in real business scenarios, improve the efficiency of test data generation, and reduce the cost of manually generating test data.
[0026] Furthermore, this method, by acquiring review process data and verifying it in conjunction with the expected review results in the test case matrix, can completely trace the status transitions, operational details, and message delivery status of the entire review process. Compared to the verification methods of single interfaces or single operations in related technologies, this method achieves comprehensive verification of the integrity, logic, and data consistency of the review process. It can more accurately identify issues such as abnormal status transitions, incomplete log records, and untimely message pushes in the review process, thus improving the comprehensiveness and accuracy of test verification. Attached Figure Description
[0027] Figure 1 This is an application environment diagram of the test method for the image and text data review function in some embodiments; Figure 2 This is a flowchart illustrating the testing method for the image and text data review function in some embodiments; Figure 3 This is a flowchart illustrating the steps involved in verifying audit results in some embodiments; Figure 4 This is a structural block diagram of a test system for the image and text data review function in some embodiments; Figure 5 This is a diagram showing the internal structure of a computer device in some embodiments. Detailed Implementation
[0028] To make the technical solutions and advantages of this disclosure clearer, the embodiments and related technical content of this disclosure will be further described in detail below with reference to the accompanying drawings and text description. It should be understood that the embodiments described below are only used to explain the technical solutions of the embodiments of this disclosure and are not intended to limit more possible implementations of this disclosure.
[0029] It should be noted that relational terms such as "first" and "second" appearing in this document are used only to distinguish things, states, or actions, and do not necessarily indicate or imply relative importance or order. The terms "including," "comprising," or any other variations thereof are used to indicate non-exclusive inclusion, and the included objects may not be limited to those listed in this document. The terms "multiple" or other variations are used to indicate that the number of objects is two or more.
[0030] It should be noted that in this application embodiment, the acquisition and processing of graphic data require user authorization and consent. When this application embodiment is applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions.
[0031] In a first aspect, embodiments of this disclosure provide a testing method for a text and image data review function. This method can be applied to, for example... Figure 1 The application environment shown is used to automate the testing of the image and text data verification function of ride-hailing platforms. This verifies the accuracy and standardization of the verification process for driver / vehicle-related documents (such as driver's licenses, vehicle registration certificates, and vehicle operating permits), ensuring the compliance of the qualifications of entities joining the ride-hailing platform. Figure 1 As shown, the application environment includes a test system 110 and a ride-hailing platform 120 under test. The two establish a communication connection via a network to achieve data interaction and command transmission. The test system 110, acting as both the test initiator and result verifier, can be deployed on a standalone server such as a physical server, cloud server, terminal device such as a PC terminal or workstation, or in a server cluster. The test system 110 possesses capabilities such as test case matrix construction, test data generation, test work order assembly, test work order sending, review flow data collection, and test result verification. The ride-hailing platform 120, acting as the test object and the main carrier of the image and text data review function, can be deployed on a standalone server, terminal device, or distributed server cluster. It can be a ride-hailing operation platform, a ride-hailing tenant management platform, etc., and possesses functions such as image and text data reception, review logic execution, review status flow, and related data recording. The ride-hailing platform under test receives test work orders sent by the test system 110, executes the image and text data review process, and generates review status data, log data, and message queue data, among other review flow data, for the test system 110 to collect and verify.
[0032] Of course, the testing method for the image and text data review function provided in this disclosure can also be applied to more scenarios not shown in the figures.
[0033] The testing method for image and text data review function is applied to Figure 1 Taking the test system 110 as an example, in some embodiments, such as Figure 2 As shown, the testing method for the image and text data review function includes steps S201 to S205 that the test system 110 can execute. The following is a detailed explanation of each step.
[0034] Step S201: Based on the image and text data of the ride-hailing platform under test, review the business logic and construct a test case matrix.
[0035] The data verification logic for images and text can be defined as the pre-set verification rules and processes for ride-hailing platforms regarding the image and text qualification documents of registered drivers and vehicles. For example, when the document type is "driver's license," the validity period and the permitted vehicle type must be verified to ensure they meet the requirements for ride-hailing drivers. If the clarity of the image and text data is below a preset threshold, the verification is deemed "failed and requires re-uploading." If the matching degree between the image and text information and the baseline data (such as driver's ID card registration information) is below a preset matching degree threshold, the information is deemed "inconsistent and requires manual review." These verification rules and processes constitute the verification business logic.
[0036] The dimensions for constructing the test case matrix can include document type, data quality, and information matching degree. Document type represents the type of qualification document for the text and image data to be reviewed, such as ID card, driver's license, vehicle registration certificate, and car insurance certificate. Data quality represents the clarity of the text and image data to be reviewed corresponding to the test case, and information matching degree represents the degree of matching between the text and image data to be reviewed and the benchmark verification data. The dimensions for constructing the test case matrix correspond to the dimensions for reviewing the text and image data.
[0037] In some embodiments, document types may include ID cards, vehicle registration certificates, and driver's licenses. Data quality includes a first quality dimension representing the identifiability of the text and image data to be reviewed, a second quality dimension representing authenticity, and a third quality dimension representing timeliness. Information matching degree includes exact match, partial match, and no match.
[0038] The test case matrix also includes the expected review result for each test case. The expected review result can be based on the ride-hailing platform's image and text review business logic, predicting the judgment result and / or business flow result that each test case should obtain after platform review. In some specific examples, the expected review result can include the final review status, such as approved, rejected, manually reviewed, or requiring supplementary materials. In other specific examples, the expected review result can include expected review status data, such as expected review flow data. In some specific examples, the review flow data corresponds to the final review status.
[0039] In some specific examples, the rows of the test case matrix represent individual test cases, and the columns of the test cases include the build dimensions and expected audit results.
[0040] In some specific scenarios, the test case matrix can be generated based on the text and image data supplementation and review function. Correspondingly, in the supplementation scenario, the construction dimension of the test case matrix can also include the current review status. The current review status can include under review, approved, rejected, and requiring supplementary materials, etc.
[0041] Specifically, based on the business logic of reviewing text and image data, a test case matrix is constructed. This can be done by combining the business logic of reviewing text and image data with the review scenarios of the ride-hailing platform, enumerating all possible values for each construction dimension, and then arranging and combining the values of each construction dimension to generate all possible test scenarios, i.e., test cases. Based on the review business logic of the platform under test, the expected result corresponding to each test case is determined, resulting in a complete test case matrix.
[0042] Step S202: Generate test data including text data and simulated document image data based on the construction dimension information of each test case in the test case matrix.
[0043] The text data can be structured information corresponding to the document information to be reviewed in the image and text data to be reviewed, such as the name, document number, permitted vehicle type and validity period of the driver's license, or the license plate number, vehicle model and registration date of the vehicle registration certificate.
[0044] Specifically, the test data built according to the construction dimension must ensure that the test data built matches the construction dimension of the corresponding test case. That is, based on the construction dimension information of the test case, text data and simulated document image data are generated respectively.
[0045] For example, if the construction dimension information of a certain test case is "vehicle license, fuzzy, partial matching", then the corresponding text data is the structured information of the vehicle license, which is consistent with the baseline verification data. The simulated image data is a fuzzy image of the vehicle license, and the information recognition of the image is lower than a preset threshold.
[0046] Step S203: Associate the text data and the simulated document image data to form a test work order, and send it to the ride-hailing platform to be tested.
[0047] Specifically, the text data and simulated document image data corresponding to the same test case are associated to form a standardized test work order that meets the review interface requirements of the ride-hailing platform under test, and then sent to the ride-hailing platform under test.
[0048] Step S204: Obtain the review and processing data of the test work order from the ride-hailing platform to be tested.
[0049] The audit process data can be the process data and result data generated by the ride-hailing platform under test during the audit process of the test work order.
[0050] The audit process data includes audit status data, log data, and message queue data.
[0051] In some specific examples, the review status data can include the status identifier and status transition record of the test ticket throughout the entire review process. The status identifier includes indicators during the review process and the final review status result. In some specific examples, the review status data can include the ticket identifier, current review status, status update time, status operator, and previous status. For example: "Ticket ID: TC-20260108-001; Current Status: Rejected; Status Update Time: 2026-01-08 10:25:30; Operator: System Automatic Review; Previous Status: Under Review". The review status data can be recorded in the business database of the ride-hailing platform under test.
[0052] Log data can be information recorded by the ride-hailing platform under test regarding the review process, such as operational behavior, input and output parameters, and exception information, used to trace the review process and locate problems. In some specific examples, log data may include operation timestamps, operation types, API call information, input parameter data, output parameter results, and exception logs.
[0053] Message queue data can include information such as message identifier, message type, recipient, message content, sending time, and sending status. For example: "Message ID: MSG-20260108-001; Message type: Review rejection notice; Recipient: Driver's mobile phone number 138XXXX1234; Message content: [XX ride-hailing platform] Your submitted vehicle registration certificate has failed the review. Reason: The vehicle registration certificate holder's name does not match the registered name. Please resubmit; Sending time: 2026-01-08 10:25:35; Delivery status: Sent."
[0054] Specifically, the review status and workflow data can be obtained by calling the corresponding interfaces or file information based on the data storage and output mechanism of the ride-hailing platform under test. In some specific examples, review status data can be obtained by calling the review result query interface of the ride-hailing platform under test, log record data can be extracted by reading the system log file of the ride-hailing platform under test, and message queue data can be subscribed to and received by connecting to the message queue middleware of the ride-hailing platform under test.
[0055] In some specific examples, the approval process data can be obtained by using a preset process state machine monitor. The process state machine monitor runs synchronously with the ride-hailing platform under test and is used to listen to the work order status change information, log information and message queue in the database associated with the ride-hailing platform under test, thereby obtaining the approval status data, log record data and message queue data.
[0056] Step S205: Verify the review process data based on the expected review results to obtain the test results of the image and text data review function of the ride-hailing platform under test.
[0057] Specifically, verifying the review process data based on the expected review results can be done by verifying the consistency between the review process data and the expected review results, thus obtaining the test results of the image and text data review function of the ride-hailing platform under test.
[0058] The testing method for the aforementioned text and image data verification function achieves a systematic review and coverage of multi-dimensional testing scenarios by constructing a test case matrix that includes core dimensions such as document type, data quality, and information matching degree. Compared to the fragmented test case design in related technologies, this method can comprehensively cover multi-dimensional combinations of scenarios with different document types, data qualities, and information matching degrees, improving the comprehensiveness of test coverage and effectively avoiding test vulnerabilities caused by scenario omissions. Based on the dimensional information of the test case construction, test data containing text data and simulated document image data is automatically generated. This method can accurately generate simulated document image data and corresponding text data that meet the requirements of different testing scenarios, ensuring the authenticity and diversity of test data. It can better simulate the data characteristics in real business scenarios, improve the efficiency of test data generation, and reduce the cost of manually generating test data.
[0059] Furthermore, this method, by acquiring review process data and verifying it in conjunction with the expected review results in the test case matrix, can completely trace the status transitions, operational details, and message delivery status of the entire review process. Compared to the verification methods of single interfaces or single operations in related technologies, this method achieves comprehensive verification of the integrity, logic, and data consistency of the review process. It can more accurately identify issues such as abnormal status transitions, incomplete log records, and untimely message pushes in the review process, thus improving the comprehensiveness and accuracy of test verification.
[0060] By automating the entire process to replace manual operations and data construction in related technologies, testing efficiency can be improved.
[0061] In some embodiments, document types include ID cards, vehicle registration certificates, and driver's licenses; data quality includes a first quality dimension representing the identifiability of the text and image data to be reviewed, a second quality dimension representing authenticity, and a third quality dimension representing timeliness; information matching degree includes complete match, incomplete match, and no match.
[0062] In some specific examples, document types may also include insurance certificates, ride-hailing driver's licenses, ride-hailing transport permits, and certificates of no criminal record. The first quality dimension of data quality can include clarity and ambiguity; the second quality dimension can include authenticity and forgery; and the third quality dimension can include validity and expiration.
[0063] In some embodiments, step S202 may include the following steps: The system calls a preset text data generation tool to generate structured text data based on the document type and information matching degree corresponding to the test case; it also calls a preset image data generation tool to generate simulated document image data based on the document type and data quality corresponding to the test case.
[0064] In some specific examples, the text data generation tool can be a text generation tool integrating the Faker library (an open-source pseudo-data generation tool), which can dynamically adjust the output results according to business rules (such as information matching degree and document type association requirements). The image data generation tool can be an image generation tool integrating OpenCV (Open Source Computer Vision Library, a cross-platform open-source computer vision library), PIL (Python Imaging Library, a mainstream open-source image processing library), GAN (Generative Adversarial Network), or Diffusion Model, which can support the generation of simulated document image data that matches the text data content based on preset document templates and data quality features. Among them, forgery features can be achieved by tampering with key information areas of the document, adding fake watermarks or seals, expiration features can be achieved by modifying the display content of the document's validity period field, and blur features can be achieved by adding Gaussian blur, reducing image resolution, or increasing noise.
[0065] Specifically, the generated structured text data meets the document type and information matching requirements corresponding to the test cases. The content of the text data corresponds to the document type, and the content of the generated text data is determined based on the document type. For example, when the document type is a motor vehicle driver's license, the text data content may include the driver's name, driver's license number, permitted vehicle types, validity period, and issuing authority; when the document type is a motor vehicle registration certificate, the text data content may include the license plate number, vehicle identification number, engine number, registration date, issuance date, and vehicle usage nature; when the document type is a ride-hailing driver's qualification certificate, the text data content may include the certificate holder's name, qualification certificate number, permitted vehicle types, validity period, and issuing region.
[0066] The accuracy of the content information in the text data corresponds to the information matching degree. When the information matching degree is a complete match, the generated text data is completely consistent with the benchmark verification data of the ride-hailing platform under test. When the information matching degree is a partial match, the core fields of the generated text data (such as driver's license number and license plate number) are consistent with the benchmark verification data, while the secondary fields (such as issuing authority and issuance date) differ from the benchmark verification data. When the information matching degree is a complete mismatch, the core fields of the generated text data completely contradict the benchmark verification data and have no consistency relationship whatsoever.
[0067] Specifically, based on the document type and data quality corresponding to the test cases, simulated document image data is generated. A pre-configured image data generation tool can be triggered to load the corresponding document template according to the document type, determine the size, field position, background style, and other information of the image to be generated, and then make targeted modifications to the basic template image according to the data quality requirements to generate simulated document image data that meets the document type and data quality requirements.
[0068] It is easy to understand that the preset text data generation tool and the preset image data generation tool can be functional modules or components pre-configured in the test data generation engine. They can be integrated into the same test data generation engine and called uniformly to realize text data generation and image data generation. They can be independent functional modules within the engine or a single functional component integrated into one.
[0069] By automatically generating structured text and simulated document images through preset text and image generation tools, the system replaces manual data construction, significantly reducing the cost of test data production and improving testing efficiency. At the same time, it can flexibly adapt to different document types, information matching degrees, and data quality requirements, covering the entire scenario of text and image verification on ride-hailing platforms, and enhancing the comprehensiveness and scalability of testing.
[0070] In some embodiments, after the test data is sent to the platform under test, the testing method for the image and text data review function may further include the following steps: The system uses a pre-set automated script or a pre-set application programming interface to simulate an auditor logging into the ride-hailing platform under test, triggering the platform's image and text data auditing function to audit the test data.
[0071] The preset automated scripts can be pre-written program scripts that can simulate human operation behaviors, used to simulate manual login, click, submit and other operations performed by auditors.
[0072] The pre-defined application programming interface (API) can be an interface used to call the review function of the ride-hailing platform under test.
[0073] By using automated scripts or APIs, a simulated auditor logs into the ride-hailing platform under test, triggers the platform's audit function, and executes the audit processing logic to conduct the audit.
[0074] By simulating the triggering of review operations through automated scripts or APIs, the entire process from sending test data to triggering the review can be automated without human intervention.
[0075] In some embodiments, the test results include state verification results, log verification results, and message queue verification results.
[0076] Among them, the status verification result is used to indicate the accuracy of the final audit status, the log verification result is used to indicate the completeness and rationality of the audit process records, and the message queue verification result is used to indicate the accuracy of the internal and external interactions of the ride-hailing platform under test.
[0077] like Figure 3 As shown, step S205 may include steps S301 to S303.
[0078] Step S301: Verify whether the audit status data is consistent with the expected audit flow status in the expected audit results, and obtain the status verification result.
[0079] Step S302: Verify whether the log data corresponds to the expected audit results, and obtain the log verification results.
[0080] Step S303: Verify whether the message queue data corresponds to the expected audit result, and obtain the message queue verification result.
[0081] By comprehensively verifying audit status data, log data, and message queue data, the comprehensiveness and accuracy of test results can be improved. If status verification fails, the root cause can be located through log / message queue verification. For example, is the status error caused by missing log records or abnormal module interaction? Even if the final status meets expectations, it is also possible to verify whether the process data is compliant, avoiding the hidden problem of correct results but incorrect processes. Verifying these three types of results can improve the accuracy and traceability of test results.
[0082] In some specific examples, the review process data may also include updated data from business data tables, i.e., field changes in the business data tables associated with the ride-hailing platform under test. When the review operation is completed, the fields in the business data tables need to be updated synchronously. Fields in the business data tables may include work order status, reviewer, review time, rejection reason, etc. Correspondingly, verifying the review process data may include checking whether the updated data content of each data table in the business data tables is consistent with the actual review process information of the work order. That is, ensuring that the data in each data table in the business data tables is accurately updated after the review operation.
[0083] In some specific examples, to verify the audit workflow data based on the expected audit results, a pre-set intelligent verification model can be used. This intelligent verification model can be a pre-set rule engine or a lightweight machine learning model. The lightweight machine learning model is built based on the audit business rules and can perform verification based on the pre-set rules. For example, when the submitted work order content is "expired vehicle registration certificate", the final audit status must be "rejected" and the rejection reason must include the keyword "expiration date"; while when the submitted work order is "information fully matched and clear", the final audit result must be "approved" and a "audit approved" notification must be triggered.
[0084] In some embodiments, the method further includes the step of generating a test report based on the test results corresponding to each test case in the test case matrix.
[0085] Test reports are used to record the test case pass rate, defect logs, and / or test work order status information in the test results.
[0086] Among them, the test case pass rate is used to measure the accuracy of the image and text data review function of the ride-hailing platform under test, and the defect log is used to record abnormal test result data, such as the test case ID where the actual review status does not match the expectation, the construction dimension information of the test case corresponding to the abnormal test result, and the review flow data content corresponding to the abnormal test result.
[0087] The status flow information of the test work order includes the status tracking record of each test work order throughout the entire review process of the ride-hailing platform under test, such as: the complete flow link of each review stage and the time consumption data of each stage.
[0088] By statistically analyzing the test results corresponding to each test case in the test case matrix and generating a test report, the overall performance and defects of the graphic data review function of the ride-hailing platform under test can be presented intuitively, thereby improving the efficiency of maintenance personnel or testers in updating and maintaining the platform.
[0089] In some embodiments, the method may further include the following steps: in response to an audit business rule change instruction, obtaining the target business rule corresponding to the change instruction; and updating the test case matrix according to the target business rule.
[0090] The instruction to change the audit business rules can be initiated by testers or platform maintenance personnel to indicate that the audit business rules have been adjusted. Specifically, the instruction can be triggered through system interface calls, configuration file updates, or manual operation interfaces.
[0091] In response to a change instruction for the audit business rules, the system retrieves the target business rule corresponding to the change instruction, i.e., the latest updated audit business rule, and updates the test cases according to the target business rule. Specifically, this may include adding, deleting, or modifying the existing test case matrix.
[0092] Based on business rule change instructions, the test case matrix is automatically updated, enabling dynamic synchronization between the test case matrix and platform review rules. This prevents test scenarios from becoming invalid due to rule iterations, ensuring the timeliness and accuracy of testing. There's no need to rebuild the test case matrix; only targeted updates are required, significantly reducing test maintenance costs after rule changes and improving testing efficiency.
[0093] It should be understood that, although Figures 2-3 The steps in the flowchart are shown sequentially according to the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Figures 2-3 Unless otherwise expressly stated herein, the steps illustrated and other steps involved in the embodiments are not subject to strict order restrictions and may be performed in other orders. Furthermore, at least some steps in the foregoing embodiments may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but may be performed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but may be performed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps.
[0094] In a second aspect, embodiments of this disclosure provide a testing system for image and text data review functions, such as... Figure 4 As shown, the testing system for the image and text data review function includes: a test management platform 410, a test data generation engine 420, an automated execution engine 430, a process status monitor 440, and a verification center 450.
[0095] The test management platform 410 is used to construct a test case matrix based on the image and text data review business logic of the ride-hailing platform under test. The construction dimensions of the test case matrix include at least document type, data quality, and information matching degree. The test case matrix also includes the expected review result corresponding to each test case. Data quality is used to indicate the clarity of the image and text data to be reviewed corresponding to the test case, and information matching degree is used to indicate the degree of matching between the image and text data to be reviewed and the benchmark verification data.
[0096] The test data generation engine 420 is used to generate test data, including text data and simulated document image data, based on the construction dimension information of each test case in the test case matrix.
[0097] The automated execution engine 430 is used to associate text data and simulated document image data to form a test work order and send it to the ride-hailing platform under test.
[0098] The process status monitor 440 is used to acquire the review process data of the test work order of the ride-hailing platform under test. The review process data includes review status data, log record data and message queue data.
[0099] Verification Center 450 is used to verify the review process data based on the expected review results, and to obtain the test results of the image and text data review function of the ride-hailing platform under test.
[0100] In some specific examples, the test management platform 410 can be equivalent to the central controller of the test system, and can be used to manage various aspects of the test process, including managing the test case matrix, configuring test plans, triggering test execution with one click, viewing test progress, and generating visual test reports.
[0101] The Test Data Generation Engine 420 can generate test data in batches.
[0102] The automated execution engine 430 can automatically complete the submission of information entry applications and the review process by calling APIs or through automated scripts or other tools to simulate human operations on the platform's front-end interface. The information entry applications can include initial information entry applications and information supplementation applications.
[0103] The process status monitor 440 is used to monitor the business database, log information and message queue of the ride-hailing platform under test in real time, and to track the work order status jump, data update and message sending status to ensure seamless connection of each link of the review process.
[0104] Verification Center 450 is used to perform result verification and automatically asserts based on the preset business verification rules in the knowledge base. In some specific examples, Verification Center 450 can also integrate a rule engine or a simple machine learning model to perform complex test logic judgments.
[0105] In some specific examples, the testing system for image and text data review functions may also include a test case knowledge base, a test report database, and other service modules.
[0106] The test case knowledge base stores the test case matrix, the expected review results for each test case, test data generation rules, and business verification rules. The test report database records the test results for each test, performs statistical analysis, and generates test reports. Other service modules may include external services relied upon by the ride-hailing platform under test, such as data dictionary rules, data entry template configuration, OCR recognition services, and file storage services. The verification center 450 is also used to verify the call relationships between the ride-hailing platform under test and other service modules, and whether the call results are correct.
[0107] In some specific examples, the testing system for image and text data verification functions can use languages such as Python / Java as the primary development language. The test management platform 410 can be developed using a web framework (such as Django, a mainstream web backend development framework in the Python ecosystem). The test data generation engine 420 can integrate the Faker library for generating text data, and an image processing library for generating simulated ID card images. The automated execution engine 430 uses Playwright (a Microsoft open-source cross-browser automated testing tool) or the Requests library (a mainstream HTTP request library in the Python ecosystem). The process status monitor 440 acquires data by monitoring the database or calling the platform's log API.
[0108] For more specific limitations regarding the testing system for the image and text data review function, please refer to the limitations of the testing method for the image and text data review function described above. The testing system for the image and text data review function can also be used to execute further steps of the testing method for the image and text data review function in the embodiments of this disclosure, which will not be repeated here. Each module in the above-described testing system for the image and text data review function can be implemented entirely or partially through software, hardware, or a combination thereof. Each module can be embedded in or independent of the processor in a computer device in hardware form, or it can be stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.
[0109] In a third aspect, embodiments of this disclosure provide a computer device including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the test method for the image and text data review function provided in any embodiment of the first aspect of this disclosure.
[0110] In some embodiments, the computer device may be a server, and its internal structure diagram may be as follows: Figure 5As shown. The computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system, computer programs, and the database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores test case matrices, test data, test results, and other data. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it implements a test method for the image and text data review function in any embodiment of this document.
[0111] Those skilled in the art will understand that Figure 5 The structures shown are merely block diagrams of some structures related to the embodiments of this disclosure and do not constitute a limitation on the computer devices to which the embodiments of this disclosure are applied. Specific computer devices may include more or fewer components than those shown in the figures, or combine certain components, or have different component arrangements.
[0112] In a fourth aspect, embodiments of this disclosure provide a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the test method for the graphic data review function provided in any embodiment of the first aspect of this disclosure.
[0113] The computer-readable storage medium may be Figure 5 The computer device shown is a non-volatile readable storage medium.
[0114] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The aforementioned computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments of this disclosure can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.
[0115] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this disclosure.
[0116] The above embodiments merely illustrate several implementation methods of this disclosure, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of protection of this disclosure. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this disclosure, and these all fall within the scope of protection of this disclosure. Therefore, the scope of protection of this disclosure should be determined by the appended claims.
Claims
1. A testing method for a text and image data review function, characterized in that, The method includes: Based on the image and text data review business logic of the ride-hailing platform under test, a test case matrix is constructed. The dimensions of the test case matrix include document type, data quality, and information matching degree. The test case matrix also includes the expected review result corresponding to each test case. The data quality is used to represent the quality of the image and text data to be reviewed corresponding to the test case. The information matching degree is used to represent the degree of matching between the image and text data to be reviewed and the benchmark verification data. Based on the construction dimension information of each test case in the test case matrix, test data including text data and simulated document image data is generated; The text data and the simulated document image data are associated to form a test work order, which is then sent to the ride-hailing platform under test. Obtain the review and processing data of the test work order from the ride-hailing platform under test. The review and processing data includes review status data, log record data, and message queue data. The review process data is verified based on the expected review results to obtain the test results of the image and text data review function of the ride-hailing platform under test.
2. The method according to claim 1, characterized in that, The document types include ID cards, vehicle registration certificates, and driver's licenses; the data quality includes a first quality dimension representing the identifiability of the text and image data to be reviewed, a second quality dimension representing authenticity, and a third quality dimension representing timeliness; the information matching degree includes complete match, incomplete match, and no match.
3. The method according to claim 1, characterized in that, The step of generating test data, including text data and simulated document image data, based on the construction dimension information of each test case in the test case matrix, includes: The preset text data generation tool is invoked to generate structured text data based on the document type and information matching degree corresponding to the test case. The preset image data generation tool is invoked to generate the simulated document image data according to the document type and data quality corresponding to the test case.
4. The method according to claim 1, characterized in that, After sending the test data to the platform under test, the method further includes: A preset automated script or a preset application programming interface is used to simulate an auditor logging into the ride-hailing platform under test, triggering the image and text data auditing function of the ride-hailing platform under test, and auditing the test data.
5. The method according to claim 1, characterized in that, The test results include status verification results, log verification results, and message queue verification results. The step of verifying the review flow data based on the expected review results to obtain the test results of the image and text data review function of the ride-hailing platform under test includes: Verify whether the audit status data is consistent with the expected audit flow status in the expected audit result to obtain the status verification result; Verify whether the log data corresponds to the expected audit result to obtain the log verification result; Verify whether the message queue data corresponds to the expected audit result to obtain the message queue verification result.
6. The method according to claim 1, characterized in that, The method further includes: Based on the test results corresponding to each test case in the test case matrix, a test report is generated. The test report is used to record the test case pass rate, defect logs, and / or test work order status flow information in the test results.
7. The method according to claim 1, characterized in that, The method further includes: In response to the audit business rule change instruction, obtain the target business rule corresponding to the change instruction; Update the test case matrix according to the target business rules.
8. A test system for image and text data verification function, characterized in that, The system includes a test management platform, a test data generation engine, an automated execution engine, a process status monitor, and a verification center; The test management platform is used to construct a test case matrix based on the image and text data review business logic of the ride-hailing platform under test. The construction dimensions of the test case matrix include document type, data quality, and information matching degree. The test case matrix also includes the expected review result corresponding to each test case. The data quality is used to indicate the clarity of the image and text data to be reviewed corresponding to the test case. The information matching degree is used to indicate the degree of matching between the image and text data to be reviewed and the benchmark verification data. The test data generation engine is used to generate test data including text data and simulated document image data based on the construction dimension information of each test case in the test case matrix. The automated execution engine is used to associate the text data and the simulated document image data to form a test work order and send it to the ride-hailing platform under test. The process status monitor is used to acquire the review and processing data of the test work order of the ride-hailing platform under test. The review and processing data includes review status data, log record data and message queue data. The verification center is used to verify the review flow data according to the expected review results, and to obtain the test results of the image and text data review function of the ride-hailing platform under test.
9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.