RPA and ai-based hazardous chemical substance auditing method, device, equipment and medium

Abstract

The application provides a dangerous chemical product auditing method and device based on RPA and AI, equipment and medium, which is applied to a dangerous chemical product export auditing system. The dangerous chemical product export auditing system comprises an RPA module. The auditing method comprises the following steps: an RPA module obtains to-be-audited data from a business system; an auditing interface is generated, and electronic single certificates and scanned files are displayed on the auditing interface; the RPA module extracts first cargo information from the electronic single certificates; the RPA module calls an AI unit to identify a preset area on the scanned files to obtain an identification result, wherein the identification result comprises second cargo information; the RPA module compares the first cargo information with the second cargo information to obtain a comparison result, obtains an auditing result for the to-be-audited data according to the comparison result, and displays the auditing result on the auditing interface. Compared with an artificial auditing process, the auditing method provided by the application improves the auditing efficiency and accuracy of the dangerous chemical product export data.

Description

Technical Field

[0001] The present invention relates to the field of RPA technology, and in particular to a hazardous chemicals audit method, device, equipment, and medium based on RPA and AI. Background Art

[0002] The export of hazardous chemicals is a key customs operation. Since the paperless customs reform, relevant documents for hazardous chemical exports are submitted through the online single window provided by the customs. Customs officers in the supervisory department receive the documents through the hazardous chemical export review system and conduct manual online review. The relevant documents include electronic documents completed by the enterprise on the terminal and PDF files scanned from offline paper documents. During the manual review process, customs officers in the supervisory department need to compare the corresponding contents in the electronic documents and PDF files one by one to ensure they are completely consistent. However, since electronic documents related to hazardous chemicals contain a large amount of chemical parameters, enterprise information, and information about the review authority, manual review is inefficient and prone to errors. Summary of the Invention

[0003] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, the present invention proposes a hazardous chemical audit method, device, equipment, and medium based on RPA and AI, which can improve the efficiency and accuracy of the audit of hazardous chemical export documents.

[0004] In a first aspect, an embodiment of the present invention provides a hazardous chemical audit method based on RPA and AI, which is applied to a hazardous chemical export audit system. The hazardous chemical export audit system includes an RPA module. The audit method includes:

[0005] The RPA module obtains the materials to be reviewed from the business system, wherein the materials to be reviewed include electronic documents and scanned documents. The electronic documents are used to indicate the electronic forms submitted by the enterprise to the hazardous chemicals export review system, and the scanned documents are used to indicate the scanned copies of the paper documents submitted by the enterprise to the hazardous chemicals export review system.

[0006] Generating a review interface, displaying the electronic document and the scanned document on the review interface;

[0007] The RPA module extracts first cargo information from the electronic document, wherein the first cargo information is used to indicate structured data recorded on the electronic form;

[0008] The RPA module calls the AI ​​unit to identify a preset area on the scanned document to obtain an identification result, where the identification result includes second cargo information, wherein the second cargo information is used to indicate unstructured data recorded on the scanned copy of the paper document, and the preset area is used to indicate an area on the scanned copy of the paper document where the unstructured data is recorded;

[0009] The RPA module compares the first cargo information and the second cargo information to obtain a comparison result, obtains an audit result for the to-be-audited document based on the comparison result, and displays the audit result on the audit interface.

[0010] In some embodiments of the present invention, there are multiple electronic documents, each of which records an application number and multiple pieces of first cargo information; a switch control is provided on the review interface, and the switch control is used to adjust the display content of the review interface; and the step of extracting the first cargo information from the electronic document by the RPA module includes:

[0011] The RPA module obtains the application number, retrieves the electronic document corresponding to the application number based on the switch control, and displays the electronic document on the review interface;

[0012] The RPA module calls the switch control to change the electronic document displayed on the review interface, displays all the first cargo information in batches on the review interface, and captures the first cargo information on the review interface.

[0013] In some embodiments of the present invention, the first cargo information includes a plurality of abbreviated information, each abbreviated information has a abbreviated identifier, and the abbreviated identifier is displayed on the review interface; the review interface is provided with a plurality of display thumbnail controls, each of which corresponds to a thumbnail identifier; and the step of capturing the first cargo information on the review interface includes:

[0014] During the process of the RPA module capturing the first cargo information, when the abbreviated identifier is captured, the RPA module calls the thumbnail display control to display the abbreviated information corresponding to the abbreviated identifier on the review interface;

[0015] The RPA module captures the thumbnail information on the review interface.

[0016] In some embodiments of the present invention, before the step of the RPA module calling the AI ​​unit to identify the preset area on the scanned document and obtaining the identification result, the method further includes:

[0017] performing annotation training on the AI ​​unit based on historical audit data, so that the AI ​​unit can locate unstructured data associated with the second cargo information from the scanned document, wherein the historical audit data has multiple categories;

[0018] The labeling training of the AI ​​unit based on historical audit data includes:

[0019] Multiple training areas are marked on the historical audit data with training numbers, and the AI ​​unit extracts graphic features and text features in each of the training areas, and performs recognition training based on the graphic features and the text features, wherein each of the training areas corresponds to one training number.

[0020] In some embodiments of the present invention, the step of the RPA module calling the AI ​​unit to identify a preset area on the scanned document and obtaining a recognition result includes:

[0021] The RPA module calls the AI ​​unit, and the AI ​​unit searches for an area to be identified on the scanned document according to the training area corresponding to the training number;

[0022] The AI ​​unit extracts the unstructured data from the training area and converts the unstructured data into structured text in the same format as the first cargo information based on natural language processing technology;

[0023] The RPA module obtains the second cargo information according to the structured text.

[0024] In some embodiments of the present invention, the step of the RPA module comparing the first cargo information and the second cargo information and obtaining the audit result of the document to be audited based on the comparison result includes:

[0025] The RPA module obtains, based on a unified field of the first cargo information and the second cargo information, the first cargo information and the second cargo information corresponding to the unified field;

[0026] The RPA module determines whether the first cargo information and the second cargo information corresponding to the unified field are the same, and obtains the audit result based on a comparison between the first cargo information and the second cargo information, wherein the unified field represents a fixed field shared by the first cargo information and the second cargo information.

[0027] In some embodiments of the present invention, the step of displaying the audit result on the audit interface includes:

[0028] When the comparison result of the first cargo information and the second cargo information indicates that they are identical, the RPA module displays an audit pass mark on the audit interface;

[0029] When the comparison result of the first cargo information and the second cargo information is characterized as different, the RPA module locates the difference information between the first cargo information and the second cargo information based on the unified field, and the difference information in a first display area of ​​the electronic document and a second display area of ​​the scanned document, displays both the electronic document and the scanned document on the review interface, and highlights the first display area and the second display area.

[0030] In a second aspect, an embodiment of the present invention provides a control device comprising at least one control processor and a memory for communicating with the at least one control processor; the memory stores instructions that can be executed by the at least one control processor, and the instructions are executed by the at least one control processor so that the at least one control processor can execute the hazardous chemical audit method based on RPA and AI as described in the first aspect above.

[0031] In a third aspect, an embodiment of the present invention provides an electronic device comprising the control device as described in the second aspect above.

[0032] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions are used to execute the hazardous chemicals audit method based on RPA and AI as described in the first aspect above.

[0033] The hazardous chemical audit method based on RPA and AI according to an embodiment of the present invention has at least the following beneficial effects: The method uses the RPA module to log in to the audit system, extract first cargo information from electronic documents and second cargo information from scanned documents, and compares the first and second cargo information to obtain an audit result based on the comparison results. The RPA and AI units complete the entire audit and comparison process, improving the efficiency and accuracy of hazardous chemical export documentation audits compared to existing manual audit processes. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 This is a flowchart of a hazardous chemicals audit method based on RPA and AI, provided by one embodiment of the present invention;

[0035] Figure 2 This is one of the schematic diagrams of the review interface provided by one embodiment of the present invention;

[0036] Figure 3 This is the second schematic diagram of the review interface provided by an embodiment of the present invention;

[0037] Figure 4 It is a structural diagram of a control device provided by another embodiment of the present invention. DETAILED DESCRIPTION

[0038] The following describes embodiments of the present invention in detail. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended only to explain the present invention and are not to be construed as limiting the present invention.

[0039] In the description of the present invention, it should be understood that descriptions involving orientations, such as up, down, front, back, left, right, etc., indicating orientations or positional relationships, are based on the orientations or positional relationships shown in the accompanying drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, they cannot be understood as limitations on the present invention.

[0040] In the description of the present invention, "several" means one or more, "many" means more than two, "greater than," "less than," and "exceed" are understood to exclude the number itself, while "above," "below," and "within" are understood to include the number itself. The use of "first" and "second" in the description is solely for the purpose of distinguishing technical features and should not be construed as indicating or implying relative importance, implicitly specifying the number of the indicated technical features, or implicitly specifying the order of the indicated technical features.

[0041] In the description of the present invention, unless otherwise clearly defined, terms such as setting, installing, and connecting should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meanings of the above terms in the present invention based on the specific content of the technical solution.

[0042] This system uses the RPA module to log into the audit system, extract the first cargo information from the electronic document, and the second cargo information from the scanned document. The first and second cargo information are then compared, and the audit results are generated based on the comparisons. The entire audit and comparison process is completed through RPA and AI units, eliminating the need for human intervention and improving the efficiency and accuracy of hazardous chemical export documentation audits.

[0043] The concepts of the present invention are explained below:

[0044] RPA: Robotic Process Automation, a software technology that automates business processes by simulating and integrating human behavior in digital systems;

[0045] AI: Artificial Intelligence refers to intelligent behavior exhibited by man-made systems. It encompasses a variety of technologies designed to enable computer systems to perform tasks that typically require human intelligence, such as learning, reasoning, problem solving, perception, understanding language, and other complex tasks.

[0046] Hazardous chemicals: These are chemicals that are toxic, corrosive, explosive, flammable, or combustion-supporting, posing a hazard to humans, facilities, and the environment. These include highly toxic chemicals and other chemicals. According to the "Catalogue of Hazardous Chemicals (2015 Edition)", hazardous chemicals are further subdivided and listed to facilitate better safety management and supervision.

[0047] The control method of the embodiment of the present invention is further described below based on the accompanying drawings.

[0048] Reference Figure 1 , Figure 1 A flowchart of a hazardous chemical export audit method based on RPA and AI is provided in an embodiment of the present invention. The method is applied to a hazardous chemical export audit system. The system includes a submission window, an RPA module, and an audit interface. The audit interface is used to display the materials to be audited. The RPA module can access the audit interface based on the auditor's system account. The hazardous chemical audit method based on RPA and AI includes but is not limited to the following steps:

[0049] In step S100, the RPA module obtains the materials to be reviewed from the business system. The materials to be reviewed include electronic documents and scanned documents. The electronic documents indicate the electronic forms submitted by the enterprise to the hazardous chemicals export review system, and the scanned documents indicate the scanned copies of the paper documents submitted by the enterprise to the hazardous chemicals export review system.

[0050] Specifically, the hazardous chemicals export review system of an embodiment of the present invention is a system for customs personnel to conduct online manual reviews on a schedule. The system provides an online submission window through which enterprises can submit materials to be reviewed to the system. After the system obtains the materials to be reviewed, it stores them in the database. Customs personnel can log in to the system's review interface through an internal account and retrieve the materials to be reviewed on the review interface.

[0051] Specifically, in the to-be-audited materials in the embodiment of the present application, the electronic single certificate is a dangerous chemical material application form filled online by the enterprise, and the dangerous chemical material export audit system can provide a form template corresponding to the dangerous chemical material export data on the submission window, the enterprise fills in the form template online, and uploads and submits to the system after completion; since the filling format is fixed through the form template, the dangerous chemical material export data on the electronic single certificate can be directly recognized by the RPA module. Further, the scanned file is a scanned copy (such as a PDF file) of a paper file offline filled by the enterprise, and after the corresponding form is filled offline and stamped at the relevant authorized agency, the paper file is scanned by a scanning tool to obtain the scanned copy, and the scanned copy is also uploaded to the system through the submission window. The characters, icons, etc. on the scanned copy such as a PDF file cannot be directly recognized. In the manual audit process, the work of the customs personnel is to check whether there are inconsistencies between the scanned copy and the electronic single certificate, and if there are inconsistencies, it is considered that the enterprise's submitted materials are incorrect, and the system notifies the enterprise to modify the materials.

[0052] It can be understood that the RPA system provided by the embodiment of the present application has the permission to log in to the system, so it can log in to the system by imitating the operation of the customs personnel logging in to the system, and call the to-be-audited materials on the audit interface as preparation for starting the audit, thereby realizing that when the enterprise submits the to-be-audited materials, the corresponding to-be-audited materials can be immediately audited, avoiding the delay of the audit process due to manual factors.

[0053] Step S200, generating an audit interface, displaying the electronic single certificate and the scanned file in the audit interface;

[0054] Step S300, the RPA module extracts first cargo information from the electronic single certificate, wherein the first cargo information is used to indicate the structured data recorded on the electronic form;

[0055] It should be noted that in steps S200-S300, the audit interface is an operation interface generated by the system based on the electronic single certificate and the scanned file. After the RPA module logs in to the system in the embodiment of the present application, it can imitate the operation process of the customs audit personnel on the audit interface, for example, by operating a series of buttons or controls on the audit interface to open the electronic single certificate or the scanned file. After the RPA module opens the electronic single certificate, since the data on the electronic single certificate is filled by the enterprise based on the form template, the data on the electronic single certificate is structured data, and the RPA module can directly recognize and obtain the first cargo information without additional data processing procedures. Further, the RPA module needs to recognize the first cargo information, including, for example, inspection date, inspection personnel, enterprise name, application number, product name, UN number, weight, value, export country (region), dangerous goods packaging use identification application number, dangerous goods packaging use identification certificate number, packaging quantity, number of cancellations, remaining quantity, etc.

[0056] In step S400, the RPA module calls the AI ​​unit to identify the preset area on the scanned document and obtain a recognition result, which includes second cargo information. The second cargo information indicates unstructured data recorded on the scanned paper document, and the preset area indicates an area on the scanned paper document where the unstructured data is recorded.

[0057] It should be noted that the embodiment of the present invention uses the AI ​​unit to identify the corresponding area on the scanned copy of the paper document, such as the PDF file, to obtain the second cargo information. Among them, the preset area records the key information that needs to be reviewed in the review of the export documents of hazardous chemicals. The types of these key information are the same as the first cargo information on the electronic document, and also include the inspection date, inspector, company name, application number, product name, UN number, weight, value of goods, export country (region), dangerous goods packaging use identification application number, dangerous goods packaging use identification certificate number, packaging quantity, verification quantity, remaining quantity and other information. It can be understood that the information obtained by the AI ​​unit from the scanned document is unstructured data with an unfixed format, and a series of data processing is required to enable this data to be recognized by the RPA module.

[0058] In step S500, the RPA module compares the first cargo information with the second cargo information to obtain a comparison result, obtains an audit result for the document to be audited based on the comparison result, and displays the audit result on the audit interface.

[0059] It should be noted that after obtaining the first cargo information and the second cargo information, the RPA module compares the corresponding parts of the first cargo information and the second cargo information, for example, simultaneously comparing all the corresponding information in the first cargo information and the second cargo information mentioned in the above-mentioned embodiments, such as the inspection date, inspector, company name, application number, etc. If all the information is the same after the comparison, it can be considered that this batch of information is qualified and the information is allowed to enter the subsequent process. If there is different information after the comparison, this batch of information is not allowed to enter the subsequent process and the information is returned to the enterprise for modification. Specifically, the reason for the error in the information is that the first cargo information does not match the second cargo information, which may be caused by the enterprise filling in the electronic document incorrectly or the process error when conducting offline qualification inspection based on paper documents. Alternatively, the reason for the error is that the second cargo information cannot be recognized, which may be because the scanned copy of the paper document is blurred and the AI ​​cannot recognize it.

[0060] It should be noted that in steps S100 to S500, this embodiment of the present invention uses the RPA module to complete the corresponding operations performed by customs personnel on the review interface during the previous manual review process, and implements the required operations performed by customs personnel on the review interface through the RPA module. Furthermore, the RPA module and the AI ​​unit it invokes can also obtain structured first cargo information from electronic documents and unstructured second cargo information from scanned documents, and then aggregate and transmit these information to the RPA module for data comparison. Compared to manual review, the speed at which the RPA module obtains the first and second cargo information is far greater than human reading speed, and the comparison is less likely to result in missed or incorrect reviews than manual review.

[0061] In some embodiments of the present invention, there are multiple electronic documents, each of which records an application number and multiple first goods information; a switch control is provided on the review interface, and the switch control is used to adjust the display content of the review interface; Figure 1 Step S200 includes but is not limited to the following steps:

[0062] In step S210, the RPA module obtains the application number, retrieves the electronic document corresponding to the application number based on the switch control, and displays the electronic document on the review interface;

[0063] In step S220, the RPA module calls a switch control to change the electronic document displayed on the review interface, displays all the first cargo information in batches on the review interface, and captures the first cargo information on the review interface.

[0064] It should be noted that in steps S210 to S220, when the RPA module retrieves the electronic document, it first needs to obtain the application number corresponding to the electronic document submitted by the enterprise from the hazardous chemicals export audit system database. It then uses a toggle control to simulate the actual operations of a customs auditor on the audit interface. For example, after obtaining the application number, the RPA module searches for it on the audit interface and simulates the auditor clicking on it. The first cargo information corresponding to the application number will pop up on the audit interface. It is understandable that electronic documents submitted by enterprises often come in multiple batches and quantities, requiring the RPA module to frequently use toggle controls to display different electronic documents on the audit interface and retrieve the first cargo information on each electronic document one by one.

[0065] In some embodiments of the present invention, the first cargo information includes a plurality of abbreviated information, each abbreviated information has a abbreviated identifier, and the abbreviated identifier is displayed on the review interface; a plurality of display abbreviated controls are provided on the review interface, each display abbreviated control corresponding to a abbreviated identifier; Figure 1 Step S200 in the illustrated embodiment also includes but is not limited to the following steps:

[0066] In step S230, when the RPA module captures the first cargo information and captures the abbreviated identifier, the RPA module calls the thumbnail display control to display the abbreviated information corresponding to the abbreviated identifier on the review interface;

[0067] In step S240 , the RPA module captures thumbnail information on the review interface.

[0068] It should be noted that in electronic documents, some of the first cargo information is very long and difficult to fully display on the review interface. Therefore, the review interface will abbreviate it to a certain extent. Customs reviewers can use the display thumbnail control on the electronic review interface to call up the abbreviated information on the review interface. Therefore, in steps S230 to S240, the RPA module also needs to call the display thumbnail control to display the abbreviated information on the review interface for subsequent capture. The RPA module determines whether there is abbreviated information based on the abbreviated mark on the electronic document. The abbreviated mark can be a different color than other first cargo information on the review interface or a button box, and there is no restriction on this.

[0069] In some embodiments of the present invention, Figure 1 Prior to step S300 in the illustrated embodiment, the following steps are included but are not limited to:

[0070] Step S301: performing annotation training on an AI unit based on historical audit data, so that the AI ​​unit can locate unstructured data associated with the second cargo information from the scanned document, wherein the historical audit data has multiple categories;

[0071] Step S302, annotating and training the AI ​​unit based on historical audit data, including: annotating multiple training areas on the historical audit data with training numbers, the AI ​​unit extracting graphic features and text features in each training area, and performing recognition training based on the graphic features and text features, wherein each training area corresponds to a training number.

[0072] It should be noted that in order for RPA to perform automated data comparison, it is necessary to extract and train the scanned copies of the paper documents submitted by the enterprise, so that the AI ​​unit can know where the second cargo information on the scanned document is located. In the embodiment of the present invention, the scanned documents that need to be extracted and compared are 8 pages per batch, including documents such as the import and export cargo packaging performance inspection result sheet, hazardous characteristics classification identification report, hazard publicity label, safety data sheet, and declaration of conformity. These documents include character segments related to the second cargo information (such as the various types of text information mentioned in the above-mentioned embodiments) and icons (hazard signs, audit agency seals, corporate seals, etc.). In order for RPA to perform automated data comparison, the AI ​​unit must first learn which areas on the scanned document need to be identified. This can be achieved by using the historical audit materials that have been reviewed and approved in documents such as the import and export cargo packaging performance inspection result sheet provided in the above-mentioned embodiments as the training set for the AI ​​unit. Further, referring to Figure 2 Before the AI ​​unit can be trained, it must process the historical audit data. For example, it can mark the areas requiring audit using markers D1-D4, D16-D19, and so on. The AI ​​then learns the training numbers for the audit data based on the markers in the historical audit data. Because the content to be recognized includes both characters and icons, the AI ​​unit must train both the graphic features and the text features within the area.

[0073] It is understandable that before the AI ​​unit recognizes the scanned documents in the embodiment of the present invention, it is necessary to first perform annotation training based on the training area annotated in the historical audit data to improve the accuracy of the AI ​​unit's recognition. If there are scanned documents in the historical audit data that deviate significantly from the rest of the data, such as scanned documents with a style that is significantly different from other scanned documents, or documents containing a mixture of Chinese and English, pure English, or other foreign languages ​​(such as English, Korean), or if there are multiple hazard notice labels in a PDF file, these documents with significant differences need to be removed from the historical audit data training set.

[0074] In some embodiments of the present invention, Figure 1 The steps of step S400 in the illustrated embodiment include but are not limited to the following steps:

[0075] In step S410, the RPA module calls the AI ​​unit, which searches for the area to be identified on the scanned document based on the training area corresponding to the training number.

[0076] In step S420, the AI ​​unit extracts unstructured data from the training area and converts the unstructured data into structured text in the same format as the first cargo information based on natural language processing technology;

[0077] In step S430 , the RPA module obtains the second cargo information according to the structured text.

[0078] It should be noted that in steps S410 to S430, when the AI ​​unit identifies the documents to be reviewed, it will directly locate the corresponding location in the training area of ​​the historical review documents required for the new batch of documents to be reviewed based on the training results of the historical review documents. Using image extraction technologies such as OCR, the second cargo information at the corresponding location in the training area is extracted. At this time, the file format of the second cargo information obtained by imaging technology is completely different from the format of the first cargo information in the electronic document and is not fixed. Therefore, before performing the comparison, it is necessary to use natural language processing technology to convert the unstructured data into structured text in the same format as the first cargo information, so that the RPA module can directly recognize the corresponding structured text.

[0079] In some embodiments of the present invention, Figure 1 The steps of step S400 in the illustrated embodiment include but are not limited to the following steps:

[0080] Step S510: The RPA module obtains the first cargo information and the second cargo information corresponding to the unified field based on the unified field of the first cargo information and the second cargo information;

[0081] In step S520, the RPA module determines whether the first cargo information and the second cargo information corresponding to the unified field are the same, and obtains an audit result based on the comparison of the first cargo information and the second cargo information. The unified field represents a fixed field shared by the first cargo information and the second cargo information.

[0082] It should be noted that in order to compare the first cargo information and the second cargo information, the RPA module needs to obtain a unified output field as the basis for comparison, that is, the first cargo information and the second cargo information associated with the same unified field are the same type of information that needs to be compared. The corresponding fields extracted by the RPA module from the electronic document or scanned document are fixed and shared by the electronic document and the scanned document. For example, Figure 3 As shown in the figure, fields such as "Requested Quantity" and "Requested Quantity Unit" appear on the form of the electronic document and the form corresponding to the paper original of the scanned document. The RPA module can compare the first cargo information and the second cargo information of the same type based on the content in the content box behind these fields.

[0083] In some embodiments of the present invention, Figure 1 The steps of step S500 in the illustrated embodiment include but are not limited to the following steps:

[0084] Step S530: When the comparison result of the first cargo information and the second cargo information indicates that they are identical, the RPA module displays an audit pass mark on the audit interface;

[0085] In step S540, when the comparison result of the first cargo information and the second cargo information indicates a difference, the RPA module locates the difference between the first cargo information and the second cargo information based on the unified field, and displays the difference in the first display area of ​​the electronic document and the second display area of ​​the scanned document. The electronic document and the scanned document are then displayed on the review interface, with the first display area and the second display area highlighted.

[0086] It should be noted that after the RPA module compares all first and second cargo information, if all of the first and second cargo information are identical, the RPA module displays a pass mark on the audit interface, allowing the batch of documents awaiting review to flow to subsequent processes. The RPA module also displays this mark on the audit interface to notify customs officers. If there are discrepancies between the first and second cargo information, such as mismatches or ambiguity, the RPA module will notify the enterprise of the failed audit through the hazardous chemicals export audit system. The RPA module will also display the specific locations where the differences between the electronic document and the scanned document occur on the audit interface for customs officers to compare and review. Specifically, the RPA module can display the audit results on the audit interface using a corresponding toggle control. The audit interface displays both the electronic document and the scanned document simultaneously, with the first display area set for the electronic document and the second display area for the scanned document. Based on the markings in the first and second display areas, customs officers can intuitively identify the discrepancies and the specific reasons for the failed audit.

[0087] like Figure 4 As shown, Figure 4 : is a structural diagram of a control device provided by an embodiment of the present invention. The present invention also provides a control device, including:

[0088] The processor 801 may be implemented using a general-purpose central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is configured to execute relevant programs to implement the technical solutions provided in the embodiments of the present application.

[0089] The memory 802 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 802 can store an operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 802 and is called by the processor 801 to execute the hazardous chemical audit method based on RPA and AI in the embodiments of this application.

[0090] Input / output interface 803, used to implement information input and output;

[0091] Communication interface 804, used to implement communication interaction between the apparatus and other devices, which can be achieved through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, Wi-Fi, Bluetooth, etc.);

[0092] Bus 805 , which transmits information between various components of the device (e.g., processor 801 , memory 802 , input / output interface 803 , and communication interface 804 );

[0093] The processor 801 , the memory 802 , the input / output interface 803 and the communication interface 804 are connected to each other in communication within the device via a bus 805 .

[0094] An embodiment of the present application further provides an electronic device, including the control device as described above.

[0095] An embodiment of the present application also provides a storage medium, which is a computer-readable storage medium. The storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned hazardous chemical audit method based on RPA and AI.

[0096] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. In addition, the memory can include a high-speed random access memory, and can also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state memory device. In some embodiments, the memory can optionally include a memory disposed remotely with respect to the processor, which can be connected to the processor through a network. Examples of the above network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof. The above-described device embodiments are only illustrative, and units described as separate components can or can not be physically separated, implemented in one place, or distributed to multiple network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the present embodiment.

[0097] Those of ordinary skill in the art can understand that all or some steps in the above disclosed method and system can be implemented as software, firmware, hardware and appropriate combinations thereof. Some or all physical components can be implemented as software executed by a processor, such as a central processing unit, a digital signal processor or a microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software can be distributed on a computer readable medium, which can include computer storage media (or non-transitory media) and communication media (or transitory media). As known to those of ordinary skill in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information such as computer readable instructions, data structures, program modules or other data. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. In addition, as known to those of ordinary skill in the art, communication media generally includes computer readable instructions, data structures, program modules or other data in modulated data signals such as carrier waves or other transmission mechanisms, and can include any information delivery medium.

[0098] The above is a specific description of the preferred embodiment of the present application, but the present application is not limited to the above-described embodiments, and those skilled in the art can make various equivalent modifications or replacements without departing from the spirit of the present application. These equivalent modifications or replacements are all included in the scope defined by the claims of the present application.

Claims

1. A hazardous chemicals audit method based on RPA and AI, characterized by: Applied to the hazardous chemicals export audit system, the hazardous chemicals export audit system includes an RPA module, and the audit method includes: The RPA module obtains the materials to be reviewed from the business system, wherein the materials to be reviewed include electronic documents and scanned documents. The electronic documents are used to indicate the electronic forms submitted by the enterprise to the hazardous chemicals export review system, and the scanned documents are used to indicate the scanned copies of the paper documents submitted by the enterprise to the hazardous chemicals export review system. Generating a review interface, displaying the electronic document and the scanned document on the review interface; The RPA module extracts first cargo information from the electronic document, wherein the first cargo information is used to indicate structured data recorded on the electronic form; The RPA module calls the AI ​​unit to identify a preset area on the scanned document to obtain an identification result, where the identification result includes second cargo information, wherein the second cargo information is used to indicate unstructured data recorded on the scanned copy of the paper document, and the preset area is used to indicate an area on the scanned copy of the paper document where the unstructured data is recorded; The RPA module compares the first cargo information and the second cargo information to obtain a comparison result, obtains an audit result for the to-be-audited document based on the comparison result, and displays the audit result on the audit interface; There are multiple electronic documents, each of which records an application number and multiple pieces of first cargo information; a switch control is provided on the review interface, and the switch control is used to adjust the display content of the review interface; and the step of extracting the first cargo information from the electronic documents by the RPA module includes: The RPA module obtains the application number, retrieves the electronic document corresponding to the application number based on the switch control, and displays the electronic document on the review interface; The RPA module calls the switch control to change the electronic document displayed on the review interface, displays all the first cargo information in batches on the review interface, and captures the first cargo information on the review interface; Before the step of the RPA module calling the AI ​​unit to identify a preset area on the scanned document and obtaining a recognition result, the method further includes: performing annotation training on the AI ​​unit based on historical audit data, so that the AI ​​unit can locate unstructured data associated with the second cargo information from the scanned document, wherein the historical audit data has multiple categories; The labeling training of the AI ​​unit based on historical audit data includes: Marking a plurality of training areas on the historical audit data with training numbers, the AI ​​unit extracting graphic features and text features in each of the training areas, and performing recognition training based on the graphic features and the text features, wherein each training area corresponds to one training number; The RPA module calls the AI ​​unit to identify the preset area on the scanned document to obtain the identification result, including: The RPA module calls the AI ​​unit, and the AI ​​unit searches for an area to be identified on the scanned document according to the training area corresponding to the training number; The AI ​​unit extracts the unstructured data from the training area and converts the unstructured data into structured text in the same format as the first cargo information based on natural language processing technology; The RPA module obtains the second cargo information according to the structured text.

2. The hazardous chemicals audit method based on RPA and AI according to claim 1 is characterized in that: The first cargo information includes a plurality of abbreviated information, each abbreviated information has a abbreviated identifier, and the abbreviated identifier is displayed on the review interface; the review interface is provided with a plurality of display thumbnail controls, each of which corresponds to a thumbnail identifier; the step of capturing the first cargo information on the review interface includes: During the process of the RPA module capturing the first cargo information, when the abbreviated identifier is captured, the RPA module calls the thumbnail display control to display the abbreviated information corresponding to the abbreviated identifier on the review interface; The RPA module captures the thumbnail information on the review interface.

3. The hazardous chemicals audit method based on RPA and AI according to claim 1 is characterized in that: The step of the RPA module comparing the first cargo information and the second cargo information and obtaining the audit result of the to-be-audited document according to the comparison result includes: The RPA module obtains, based on a unified field of the first cargo information and the second cargo information, the first cargo information and the second cargo information corresponding to the unified field; The RPA module determines whether the first cargo information and the second cargo information corresponding to the unified field are the same, and obtains the audit result based on a comparison between the first cargo information and the second cargo information, wherein the unified field represents a fixed field shared by the first cargo information and the second cargo information.

4. The RPA and AI-based hazardous chemicals audit method according to claim 3 is characterized in that: The step of displaying the audit result on the audit interface includes: When the comparison result of the first cargo information and the second cargo information indicates that they are identical, the RPA module displays an audit pass mark on the audit interface; When the comparison result of the first cargo information and the second cargo information is characterized as different, the RPA module locates the difference information between the first cargo information and the second cargo information based on the unified field, and the difference information in a first display area of ​​the electronic document and a second display area of ​​the scanned document, displays both the electronic document and the scanned document on the review interface, and highlights the first display area and the second display area.

5. A control device, characterized in that: It includes at least one control processor and a memory for communicating with the at least one control processor; the memory stores instructions that can be executed by the at least one control processor, and the instructions are executed by the at least one control processor to enable the at least one control processor to execute the hazardous chemicals audit method based on RPA and AI as described in any one of claims 1 to 4.

6. An electronic device, characterized in that: Includes the control device according to claim 5.

7. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions, which are used to enable a computer to execute the hazardous chemical audit method based on RPA and AI as described in any one of claims 1 to 4.

Images