Source data submission method and device based on artificial intelligence, and electronic equipment

Abstract

The invention discloses a source data submission method and device based on artificial intelligence and electronic equipment, and relates to the field of artificial intelligence or financial science and technology, and the method comprises the steps: carrying out the character recognition of a submitted data standard file through an optical character recognition algorithm, and when a target character or a thickened character exists in a recognition result, carrying out the recognition of a submitted data standard file; and determining the region of the target character or the thickened character as a change point, extracting a header and a table name of a report where the change point is located and all information of a line to which the change point belongs, determining the header, the table name and the information of the line to which the change point belongs as change content of the change point, and obtaining original transaction data. And modifying the submitted data generation script based on the change point and the corresponding change content, generating submitted source data, performing format conversion and packaging processing on the submitted source data, and outputting the packaged submitted source data. According to the method and the device, the technical problem of low submission efficiency caused by the fact that a source data submission mode depends on manual submission and submission system transformation in the prior art is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of artificial intelligence, the field of financial technology or other related fields, in particular, to a source data reporting method and device based on artificial intelligence, and an electronic device. BACKGROUND

[0002] In the financial market, the data reporting of the financial planning subsidiary is an important part of financial regulation. Through regular data reporting, the regulatory authorities can timely grasp the asset position, business activities and risk level of financial institutions, and thus ensure the stability and transparency of the financial market. With the increasing complexity and diversity of financial business, the frequency and specifications of data reporting are constantly changing. Therefore, the financial planning subsidiary needs to regularly update the generation mechanism of its reported data to adapt to the latest requirements of the regulatory authorities and ensure the accuracy and timeliness of the reported data.

[0003] In related technologies, in the scenario of constantly changing reporting specifications, the existing financial data reporting strategy has significant shortcomings. When facing the scenario of frequent changes in reporting data specifications, the financial planning subsidiary usually relies on manual sorting of new reporting requirements by business personnel, then submits to the IT department to modify the reporting system, and then uses the modified reporting system to generate reporting source data, and finally carries the reporting source data to the regulatory authorities by the business personnel. This process is complex and time-consuming, not only prone to human error, but also causes delays in data reporting, which cannot meet the requirements of the regulatory authorities for the timeliness and accuracy of data reporting. In addition, this strategy relying on manual reporting and reporting system modification requires repeated investment of research and development resources when facing frequent changes in reporting specifications, greatly increasing the operating costs of the financial planning subsidiary.

[0004] In view of the above problems, no effective solutions have been proposed so far. SUMMARY

[0005] The embodiments of the present application provide a source data reporting method and device based on artificial intelligence, and an electronic device, to at least solve the technical problem of low reporting efficiency caused by the dependence of the related art source data reporting method on manual reporting and reporting system modification.

[0006] In order to achieve the above object, according to one aspect of the present application, a source data reporting method based on artificial intelligence is provided, comprising: detecting a reporting data specification file, and performing character recognition on the reporting data specification file by using an optical character recognition algorithm, wherein the content in the reporting data specification file is in the form of a report; in the case that there is a target character or a bold character in the recognition result, determining the region of the target character or the bold character as a change point, wherein the change point includes renaming, adding, adjusting and reducing; for each change point, extracting the table header and the table name of the report in which the change point is located and all information of the row to which the change point belongs, and determining the table header, the table name and all information of the row to which the change point belongs as the change content of the change point; obtaining original transaction data, modifying a pre-constructed reporting data generation script based on the change point and the corresponding change content, and generating reporting source data based on the modified reporting data generation script and the original transaction data; performing format conversion processing and packaging processing on the reporting source data, and outputting the reporting source data after packaging.

[0007] Optionally, the step of performing character recognition on the reporting data specification file by using an optical character recognition algorithm comprises: performing positioning processing on the reporting data specification file by using a selective search algorithm to generate a plurality of recognition candidate regions; inputting the reporting data specification file into a pre-constructed deep network to obtain a reporting feature map; projecting each recognition candidate region to the reporting feature map to generate a plurality of feature matrices, and inputting each feature matrix into a pooling layer to obtain a plurality of feature matrices of uniform size; inputting each feature matrix into a fully connected layer and an activation layer to obtain a plurality of recognition characters, wherein the recognition characters include target characters and bold characters.

[0008] Optionally, after determining the change point and the corresponding change content of the target character or the bold character, comprising: determining the change point and the corresponding change content of the target character as a target character change set, and determining the change point and the corresponding change content of the bold character as a bold character change set; comparing the target character change set and the bold character change set, and taking the union as a target change set, wherein the target change set is used to modify the pre-constructed reporting data generation script.

[0009] Optionally, in the case that the change point is renaming, the step of generating reporting source data comprises: determining the corresponding item of the change point as a target item, modifying the name of the corresponding item in the reporting data generation script based on the serial number and the name of the target item in the change content; screening out the transaction data belonging to the target item from the original transaction data, and generating reporting source data based on the modified item name in the reporting data generation script.

[0010] Optionally, when the change point is an addition, the step of generating the submission source data comprises: determining a project corresponding to the change point as a target project, modifying a statistical range of the target project in the submission data generation script based on a name of the target project in the change content and the statistical range description information; screening transaction data belonging to the modified statistical range from the original transaction data, and performing operation processing on the screened transaction data based on operation logic specified in the submission data generation script to generate the submission source data.

[0011] Optionally, when the change point is an adjustment, the step of generating the submission source data comprises: determining a project corresponding to the change point as a target project, and determining all sub-projects of the target project; modifying an index level of the target project in the submission data generation script based on a name of the target project in the change content and the index level, wherein the target project and all sub-projects are included in the corresponding project, and the index level is used to describe a classification level of the target project in the submission system; screening transaction data belonging to the target project and the sub-projects from the original transaction data, and generating the submission source data based on the modified index level in the submission data generation script.

[0012] Optionally, when the change point is a reduction, the step of generating the submission source data comprises: determining a project corresponding to the change point as a target project, modifying operation logic of the target project in the submission data generation script based on a name of the target project in the change content and the index level; screening transaction data of the target project from the original transaction data, and performing operation processing on the screened transaction data based on the modified operation logic in the submission data generation script to generate the submission source data.

[0013] According to another aspect of the embodiment of the present application, a source data submission device based on artificial intelligence is also provided, comprising: an identification unit configured to detect a submission data specification file, and perform character recognition on the submission data specification file by using an optical character recognition algorithm, wherein the content in the submission data specification file is presented in the form of a report; a first determination unit configured to determine a region of a target character or a bold character as a change point in the case that the target character or the bold character exists in the recognition result, wherein the change point comprises: renaming, adding, adjusting, and reducing; a second determination unit configured to, for each change point, extract a table header and a table name of a report in which the change point is located and all information of a row to which the change point belongs, and determine the table header, the table name, and all information of the row as change content of the change point; a generation unit configured to obtain original transaction data, modify a pre-constructed submission data generation script based on the change point and the corresponding change content, and generate submission source data based on the modified submission data generation script and the original transaction data; and an output unit configured to perform format conversion processing and packaging processing on the submission source data, and output the submission source data after being packaged.

[0014] Optionally, the recognition unit includes: a first processing module for locating the submitted data specification file using a selective search algorithm to generate multiple recognition candidate regions; a second processing module for inputting the submitted data specification file into a pre-constructed deep network to obtain a submitted feature map; a third processing module for projecting each recognition candidate region onto the submitted feature map to generate multiple feature matrices, and inputting each feature matrix into a pooling layer to obtain multiple feature matrices of uniform size; and a fourth processing module for inputting each feature matrix into a fully connected layer and an activation layer to obtain multiple recognition characters, wherein the recognition characters include: target characters and bold characters.

[0015] Optionally, the second determining unit includes: a first determining module, used to determine the change points of the target character and the corresponding change content as the target character change set, and to determine the change points of the bold character and the corresponding change content as the bold character change set; and a comparison module, used to compare the target character change set with the bold character change set, and take the union as the target change set, wherein the target change set is used to modify the pre-built reporting data generation script.

[0016] Optionally, the generation unit includes: a first modification module, used to determine the project corresponding to the change point as the target project when the change point is a name change, and modify the name of the corresponding project in the reporting data generation script based on the serial number and name of the target project in the change content; and a first generation module, used to filter out the transaction data belonging to the target project from the original transaction data, and generate the reporting source data based on the modified project name in the reporting data generation script.

[0017] Optionally, the generation unit further includes: a second modification module, used to determine the project corresponding to the change point as the target project when the change point is newly added, and modify the statistical range of the corresponding project in the reporting data generation script based on the name and statistical range description information of the target project in the change content; and a second generation module, used to filter out transaction data that belongs to the modified statistical range from the original transaction data, and perform calculation processing on the filtered transaction data based on the calculation logic specified in the reporting data generation script to generate reporting source data.

[0018] Optionally, the generation unit further includes: a second determining module, used to determine the project corresponding to the change point as the target project when the change point is an adjustment, and to determine all subordinate projects of the target project; a third modifying module, used to modify the indicator level of the corresponding project in the reporting data generation script based on the name and indicator level of the target project in the change content, wherein the corresponding project includes: the target project and all subordinate projects, and the indicator level is used to describe the classification level of the target project in the reporting system; and a third generation module, used to filter out the transaction data belonging to the target project and subordinate projects from the original transaction data, and generate the reporting source data based on the modified indicator level in the reporting data generation script.

[0019] Optionally, the generation unit further includes: a fourth modification module, used to determine the project corresponding to the change point as the target project when the change point is narrowed, and modify the calculation logic of the target project in the reporting data generation script based on the name and indicator level of the target project in the change content; and a fourth generation module, used to filter out the transaction data of the target project from the original transaction data, and perform calculation processing on the filtered transaction data based on the modified calculation logic of the reporting data generation script to generate the reporting source data.

[0020] According to another aspect of the present invention, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform any of the above-described artificial intelligence-based source data reporting methods.

[0021] According to another aspect of the present invention, an electronic device is also provided, including one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the artificial intelligence-based source data reporting method described above.

[0022] According to another aspect of the present invention, a computer program product is also provided, including a computer program that, when executed by a processor, implements the steps of the artificial intelligence-based source data reporting method described in any one of the above embodiments.

[0023] In this disclosure, by detecting the data submission specification document, an optical character recognition (OCR) algorithm is used to perform text recognition on the document. If target characters or bolded characters are found in the recognition results, the area containing these characters is identified as a change point. For each change point, the header and name of the report containing the change point, as well as all information of the row to which the change point belongs, are extracted. The header, name, and row information are then identified as the change content for that change point. The original transaction data is then obtained, and the pre-built data submission generation script is modified based on the change points and their corresponding change content. Based on the modified reporting data generation script and the original transaction data, the system generates reporting source data. Finally, it performs format conversion and packaging of the reporting source data and outputs the packaged reporting source data. This system achieves intelligent parsing of reporting data specification files, automatic location of change points, intelligent modification of reporting data generation scripts, and automatic generation and sending of reporting source data. This greatly enhances the automation level and efficiency of source data reporting, and solves the technical problem of low reporting efficiency caused by reliance on manual reporting and reporting system modifications in related technologies. Attached Figure Description

[0024] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0025] Figure 1 A hardware structure block diagram of a computer terminal (or mobile device) for implementing an artificial intelligence-based source data reporting method is shown.

[0026] Figure 2 This is a flowchart of an optional source data reporting method based on artificial intelligence according to an embodiment of the present invention;

[0027] Figure 3 This is a simplified flowchart of an optional source data reporting method based on artificial intelligence according to an embodiment of the present invention;

[0028] Figure 4 This is a schematic diagram of an optional artificial intelligence-based source data reporting device according to an embodiment of the present invention;

[0029] Figure 5 This is a structural block diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0030] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0031] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0032] It should be noted that the source data reporting method and apparatus based on artificial intelligence in this disclosure can be used in the fields of artificial intelligence and fintech for source data reporting based on artificial intelligence, and can also be used in any field other than artificial intelligence and fintech for source data reporting based on artificial intelligence. This disclosure does not limit the application field of the source data reporting method and apparatus based on artificial intelligence.

[0033] It should be noted that the information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, and displayed data) collected in this public disclosure are information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of related data all comply with the relevant laws, regulations, and standards of the relevant regions, necessary confidentiality measures have been taken, and they do not violate public order and good morals. Corresponding operation entry points are provided for users to choose to authorize or refuse. For example, this system has interfaces with relevant users or organizations. Before obtaining relevant information, a request to obtain the information needs to be sent to the aforementioned user or organization through the interface, and the relevant information is obtained only after receiving consent from the aforementioned user or organization.

[0034] It should be noted that in this disclosure, customer information is collected and analyzed, and users are provided with corresponding operation entry points to choose whether to agree to or reject the automated decision results; if the user chooses to reject, the process will proceed to the expert decision-making process.

[0035] The following embodiments of the present invention can be applied to various AI-based source data reporting systems / applications / devices. The present invention uses an optical character recognition algorithm to perform text recognition on the reporting data specification document. When target characters or bold characters are present in the recognition results, the area containing the target characters or bold characters is identified as a change point. The header, table name, and all information of the row containing the change point are then identified as the change content for that change point. Original transaction data is then acquired, and a pre-built reporting data generation script is modified based on the change point and the corresponding change content. Reporting source data is generated based on the modified reporting data generation script and the original transaction data. Finally, the reporting source data undergoes format conversion and packaging, and the packaged reporting source data is output. This achieves intelligent parsing of the reporting data specification document, automatic location of change points, intelligent modification of the reporting data generation script, and automatic generation and transmission of reporting source data, greatly enhancing the automation and efficiency of source data reporting.

[0036] The present invention will now be described in detail with reference to various embodiments.

[0037] Example 1

[0038] According to an embodiment of the present invention, an embodiment of a source data reporting method based on artificial intelligence is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0039] The source data reporting method based on artificial intelligence provided in Embodiment 1 of this application can be executed on a mobile terminal, computer terminal, or similar computing device. Figure 1 A hardware block diagram of a computer terminal (or mobile device) for implementing an artificial intelligence-based source data reporting method is shown. Figure 1 As shown, computer terminal 10 (or mobile device) may include one or more ( Figure 1(Illustrated as 102a, 102b, ..., 102n) Processor 102 (processor 102 may include, but is not limited to, a microprocessor MCU (Microcontroller Unit) or a programmable gate array (FPGA), etc.), memory 104 for storing data, and transmission device 106 for communication functions. In addition, it may include: a display, input / output interface (I / O interface), Universal Serial Bus (USB) port (which may be included as one of the ports of a BUS bus), network interface, power supply, and / or camera. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0040] It should be noted that the aforementioned one or more processors 102 and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the computer terminal 10 (or mobile device). As involved in the embodiments of this application, the data processing circuits serve as a processor control mechanism (e.g., selection of a variable resistor termination path connected to an interface).

[0041] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the source data reporting method based on artificial intelligence in this embodiment. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the aforementioned source data reporting method based on artificial intelligence. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0042] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0043] The display can be, for example, a touchscreen liquid crystal display (LCD), which allows the user to interact with the user interface of the computer terminal 10 (or mobile device).

[0044] Under the aforementioned operating environment, this application provides the following: Figure 2 The method for reporting source data based on artificial intelligence is shown. Figure 2 This is a flowchart of an optional source data reporting method based on artificial intelligence according to an embodiment of the present invention, such as... Figure 2 As shown, the method includes the following steps:

[0045] Step S201: Detect the data submission specification document and use an optical character recognition algorithm to perform text recognition on the data submission specification document. The content of the data submission specification document is presented in the form of a report.

[0046] In financial regulation, wealth management subsidiaries are required to regularly submit financial data to regulatory agencies to ensure business compliance and market transparency. These data submission guidelines are published in document form, typically in report format, detailing the submitted indicators, field names, data formats, and statistical rules. This invention first obtains the latest version of the data submission guidelines from the regulatory agency's website, or can periodically and automatically search the agency's website for new guidelines, automatically downloading and obtaining the latest version if found.

[0047] After obtaining the latest version of the reporting data specification file, this embodiment of the invention employs an optical character recognition (OCR) algorithm to perform deep analysis and text recognition on the reporting data specification file. Optical character recognition (OCR) is an artificial intelligence algorithm that converts characters on an image into a text-formatted character sequence.

[0048] The data submission specification document exists in PDF or other image formats. In this embodiment of the invention, an optical character recognition algorithm is used to scan each page of the data submission specification document to identify all text information, including data and descriptive text in the report, in order to further capture the change points and changes in the data submission specification document.

[0049] Step S202: If the target character or bold character exists in the recognition result, the area of ​​the target character or bold character is determined as a change point. The change points include: renaming, adding, adjusting and shrinking.

[0050] After the optical character recognition algorithm performs text recognition on the submitted data specification document, it generates detailed text recognition results. In these results, when specific target characters or bold text are present, the areas containing these target characters or bold text are identified as change points.

[0051] Target characters typically refer to keywords used to mark changes in the data submission specification file, including: rename, add, adjust, and reduce. The appearance of these keywords signifies a change in the report content, and the types of changes include: rename, add, adjust, and reduce. Similarly, the appearance of bold text also signifies a change in the report content, and the types of changes also include: rename, add, adjust, and reduce. That is, the types of change points in this embodiment of the invention include: rename, add, adjust, and reduce.

[0052] Step S203: For each change point, extract the header and name of the report where the change point is located, as well as all information of the row to which the change point belongs, and determine the header, name, and all information of the row to which the change point belongs as the change content of that change point.

[0053] After identifying all change points in the data submission specification document, this embodiment of the invention further analyzes the text recognition results. First, this embodiment of the invention extracts the header and name of the report containing the change point. The header contains the column headings and key indicators of the report, while the name contains the report's name and number.

[0054] Secondly, this embodiment of the invention extracts all information from the row to which the change point belongs. Since the content of the data submission specification file exists in the form of a report, when extracting all information from the row to which the change point belongs, information from an entire row of the report is extracted, such as in Table 1:

[0055] Table 1

[0056]

[0057] Table 1 shows a portion of the reports in the data submission specification document. As shown in Table 1, "Added in 2024" is a change point. "Column number", "name", "length", "whether it can be nullable", "description", and "remarks" are all part of the table header. All information in the row to which the change point belongs includes: "3", "product type", "1c", "N", "Added in 2024" and "1-Non-principal-guaranteed wealth management products of financial institutions 2-Trust products of trust companies 3-Asset management products of securities companies and their subsidiaries 4-Special accounts of fund management companies and their subsidiaries 5-Asset management products of futures companies and their subsidiaries 6-Insurance asset management products 7-Asset management products of financial asset investment companies 8-Public funds".

[0058] After extracting the header, table name, and all information of the row to which the change point belongs from the report where the change point is located, the header, table name, and all information of the row to which the change point belongs are determined as the change content of the change point. All the change points and corresponding change content obtained will be used in the subsequent modification operation of the report data generation script.

[0059] Step S204: Obtain the original transaction data, modify the pre-built reporting data generation script based on the change points and corresponding change content, and generate the reporting source data based on the modified reporting data generation script and the original transaction data.

[0060] In this embodiment of the invention, raw transaction data is obtained from the underlying database of the wealth management subsidiary. This raw transaction data forms the basis for generating the reporting source data. The underlying database is a centralized storage platform within the wealth management subsidiary that stores all business data, covering a full range of business data, including product issuance data, customer subscription and redemption data, product termination data, and investment transaction data.

[0061] After obtaining the raw transaction data, the pre-identified and extracted change points and corresponding change content are used to automatically adjust and modify the pre-built reporting data generation script. It should be noted that the pre-built reporting data generation script is a historical one. This embodiment of the invention modifies the historical reporting data generation script based on the change points and corresponding change content to generate a reporting data generation script that conforms to the latest reporting specifications.

[0062] After the data generation script is modified, this embodiment of the invention will automatically execute the script to accurately filter and calculate the original transaction data, generating source data that conforms to the latest reporting specifications. It should be noted that the data generation script includes modified and unmodified parts. The unmodified part will generate source data according to the original script logic, while the modified part will generate source data according to the new script logic. This process is fully automated, requiring no manual intervention, and greatly improves the efficiency of source data generation while ensuring high accuracy.

[0063] Step S205: Perform format conversion and packaging processing on the source data to be reported, and output the packaged source data.

[0064] Because regulatory agencies require source data to be submitted in specific formats, such as converting source data in Excel spreadsheet format to .dat or .txt format files, the source data needs to be formatted after it is generated. Simultaneously, the source data needs to be packaged, merging multiple processed source data files into a unified source data package.

[0065] Finally, the embodiment of the present invention outputs the source data, after format conversion and packaging, and sends it to the regulatory agency via a direct connection. The direct connection method refers to sending the source data through a dedicated communication line between the wealth management subsidiary and the regulatory agency. Sending the source data via a direct connection significantly improves transmission efficiency and security.

[0066] After receiving and analyzing the submitted source data, the regulatory agency will return the submission result to the wealth management subsidiary, such as submission success, submission failure, and the reason for failure.

[0067] It should be noted that the embodiments of the present invention also support manual operation of the reporting source data. That is, when the reporting source data fails to be reported to the regulatory agency, the external terminal or the object of operation can regenerate the reporting source data or manually modify the reporting source data according to the reason for the failure, and then resubmit the verified reporting source data.

[0068] Optionally, the step of using an optical character recognition algorithm to perform text recognition on the submitted data specification document includes: using a selective search algorithm to locate the submitted data specification document and generate multiple recognition candidate regions; inputting the submitted data specification document into a pre-constructed deep network to obtain a submitted feature map; projecting each recognition candidate region onto the submitted feature map to generate multiple feature matrices, and inputting each feature matrix into a pooling layer to obtain multiple feature matrices of uniform size; inputting each feature matrix into a fully connected layer and an activation layer to obtain multiple recognized characters, wherein the recognized characters include: target characters and bold characters.

[0069] In this embodiment of the invention, an optical character recognition (OCR) algorithm is used to perform text recognition on the submitted data specification document. The main steps of the OCR algorithm include image extraction, text detection, and text recognition. The OCR algorithm used in this embodiment of the invention is based on the FastRCNN (Fast Region Convolutional Neural Network) model to perform text recognition on the submitted data specification document. The FastRCNN model is a deep learning model for object detection. It uses a Region Proposal Network (RPN) to find regions in an image where objects may exist, and then uses a convolutional network to extract features and classify these regions.

[0070] In optical character recognition algorithms, the FastR-CNN model is used to locate and recognize text content in images. It can recognize various forms of text, including printed, handwritten, and even unstructured text. The FastR-CNN model is a two-stage task model; it first locates the text region and then performs character recognition, making the model highly efficient and accurate in handling text recognition tasks in complex scenes.

[0071] When using optical character recognition (OCR) algorithms to perform text recognition on submitted data specification documents, a selective search algorithm is first used to locate the data specification documents. The selective search algorithm is an image segmentation algorithm used for object detection. In this embodiment of the invention, the submitted data specification document is first read and converted into a processable digital image. Then, the selective search algorithm automatically analyzes the image structure of the document, identifies multiple regions containing text information, and marks them as potential candidate recognition regions.

[0072] In this embodiment of the invention, the image of the reporting data specification file is input into a pre-constructed deep network to obtain a reporting feature map. Then, all candidate recognition regions are projected onto the reporting feature map to generate multiple feature matrices. In this embodiment, the deep network uses a VGG16 network, a deep learning network widely used in image recognition. Processing the image of the reporting data specification file using VGG16 generates a reporting feature map containing key features of the file. Finally, all candidate recognition regions are projected onto the reporting feature map to obtain feature matrices containing key features of the candidate recognition regions.

[0073] Then, all feature matrices are fed into a pooling layer, which scales them to the same size, for example, a uniform 7×7 feature matrix. The purpose of the pooling layer is to reduce the dimensionality of the feature matrices while preserving their key information.

[0074] Finally, the uniformly sized feature matrix is ​​sequentially input into the fully connected layer and the activation layer for character recognition processing, ultimately outputting multiple recognized characters, including the target character and bold characters. This embodiment of the invention, by employing an optical character recognition algorithm combined with a selective search algorithm, deep network processing, feature matrix generation, and character recognition, achieves intelligent identification and location of change points in the reported data specification file, providing accurate change information support for subsequent modification of the reported data generation script and generation of the reported source data.

[0075] It should be noted that the optical character recognition (OCR) algorithm was trained and optimized using historical data beforehand. This historical data includes samples of previous data submission specifications. These samples not only cover different submission requirements and formats but also contain various types of change records, such as field name renaming, type additions, and statistical range reductions. By collecting and organizing this historical data, the OCR algorithm is trained and continuously optimized using a comprehensive and diverse historical dataset. This ensures that the algorithm can accurately identify and extract key information from the data submission specifications, especially information related to changes. This provides precise and complete data support for subsequent modifications to the data submission generation script and the generation of source data.

[0076] Optionally, after determining the change points and corresponding change content of the target character or bold character, the process includes: determining the change points and corresponding change content of the target character as the target character change set, and determining the change points and corresponding change content of the bold character as the bold character change set; comparing the target character change set with the bold character change set, and taking the union as the target change set, wherein the target change set is used to modify the pre-built reporting data generation script.

[0077] In this embodiment of the invention, the character types that will be identified as change points include two types: target characters and bold characters. However, there is a phenomenon of repetition in the change points determined by identifying target characters or bold characters, that is, a certain change point will be identified as both target characters and bold characters at the same time.

[0078] In this embodiment of the invention, the change points and corresponding change content of the target character are determined as the target character change set, and the change points and corresponding change content of the bold character are determined as the bold character change set. Then, the target character change set and the bold character change set are compared, and the union is taken as the target change set. That is, when the change points and corresponding change content of the target character are exactly the same as those of the bold character, only the change points and corresponding change content of one of them are retained as subsequent change information.

[0079] Optionally, when the change point is a name change, the steps for generating reporting source data include: identifying the project corresponding to the change point as the target project, modifying the name of the corresponding project in the reporting data generation script based on the sequence number and name of the target project in the change content; filtering out transaction data belonging to the target project from the original transaction data, and generating reporting source data based on the modified project name in the reporting data generation script.

[0080] When the change point is a name change, see Table 2 below:

[0081] Table 2

[0082]

[0083] Table 2 shows some of the reports in the data submission specification document. As shown in Table 2, "2024 name change" is identified as a change point of the name change type. The table header includes: "line number", "item", "RMB", "USD to RMB", "EUR to RMB", "GBP to RMB", "JPY to RMB", "other currencies to RMB", "total" and "remarks". The change content includes: "811200", "repurchase and interbank lending (including loans)" and "2024 name change".

[0084] When generating the source data for reporting, the project corresponding to the change point is first identified as the target project, specifically "Repurchase and Lending (including Loans)". Then, based on the sequence number and name of the target project in the changed content, the name of the corresponding project in the reporting data generation script is modified. This means changing the name of the corresponding project in the reporting data generation script according to the row number or column number of the row in the changed content, as well as the project name. For example, if the original reporting data generation script named the target project "Repurchase and Lending", after renaming, the modified reporting data generation script will name the target project "Repurchase and Lending (including Loans)".

[0085] After the name of the target project is modified, this embodiment of the invention filters out transaction data belonging to the target project from the original transaction data, and then generates reporting source data based on the modified project name in the reporting data generation script. For example, transaction data belonging to "repurchase and interbank lending (including loans)" is filtered out from the original transaction data, and arithmetic operations such as addition, subtraction, multiplication, and division are performed on the transaction data according to the calculation logic specified in the reporting data generation script, finally generating reporting source data of the "repurchase and interbank lending (including loans)" category.

[0086] Optionally, when the change point is newly added, the steps for generating the reporting source data include: determining the project corresponding to the change point as the target project; modifying the statistical range of the corresponding project in the reporting data generation script based on the name and statistical range description information of the target project in the change content; filtering out transaction data that falls within the modified statistical range from the original transaction data; and performing calculations on the filtered transaction data based on the calculation logic specified in the reporting data generation script to generate the reporting source data.

[0087] When a change point is newly added, as shown in Table 1, "New in 2024" is identified as a new type of change point. When generating the source data for reporting, the project corresponding to the change point is first identified as the target project, that is, "Product Type" is identified as the target project. Then, based on the name and statistical scope description information of the target project in the change content, the statistical scope of the corresponding project in the reporting data generation script is modified. That is, according to the name "Product Type" and the statistical scope description information "1-Non-principal-protected wealth management products of financial institutions 2-Trust products of trust companies 3-Asset management products of securities companies and their subsidiaries 4-Special accounts of fund management companies and their subsidiaries 5-Asset management products of futures companies and their subsidiaries 6-Insurance asset management products 7-Asset management products of financial asset investment companies 8-Public funds", the statistical scope of the corresponding project in the reporting data generation script is modified.

[0088] Then, transaction data that falls within the modified statistical range is selected from the original transaction data. Specifically, transaction data belonging to the following eight product types are selected from the original transaction data: "1- Financial institutions' non-principal-protected wealth management products; 2- Trust products of trust companies; 3- Asset management products of securities companies and their subsidiaries; 4- Separate accounts of fund management companies and their subsidiaries; 5- Asset management products of futures companies and their subsidiaries; 6- Insurance asset management products; 7- Asset management products of financial asset investment companies; 8- Public funds". Based on the calculation logic specified in the reporting data generation script, the selected transaction data is processed to generate the reporting source data for these eight product types.

[0089] Optionally, when the change point is an adjustment, the step of generating the reporting source data includes: identifying the project corresponding to the change point as the target project and identifying all subordinate projects of the target project; modifying the indicator level of the corresponding project in the reporting data generation script based on the name and indicator level of the target project in the change content, wherein the corresponding project includes: the target project and all subordinate projects, and the indicator level is used to describe the classification level of the target project in the reporting system; filtering out the transaction data belonging to the target project and subordinate projects from the original transaction data, and generating the reporting source data based on the modified indicator level in the reporting data generation script.

[0090] When the change point is adjustment, see Table 3 below:

[0091] Table 3

[0092]

[0093] Table 3 shows some reports in the data submission specification document. As shown in Table 3, "2024 Adjustment Indicator Level" is identified as a change point of adjustment type. The table header includes: "Line Number", "Item", "RMB", "USD to RMB", "EUR to RMB", "GBP to RMB", "JPY to RMB", "Other Currencies to RMB", "Total" and "Remarks". The changed content includes: "7d1000", "Asset Income Rights" and "2024 Adjustment Indicator Level".

[0094] When generating source data for reporting, the project corresponding to the change point is first identified as the target project, i.e., "Asset Income Rights" is identified as the target project. Then, based on the name and indicator level of the target project in the change content, the indicator level of the corresponding project in the reporting data generation script is modified. In this process, the corresponding project includes the target project and all subordinate projects, and the indicator level refers to the line number. For example, if the indicator level of "Asset Income Rights" is adjusted from the first level to the second level, i.e., the line number "7d0000" is adjusted to "7d1000", the line number of "Asset Income Rights" in the reporting data generation script is modified to "7d1000" according to this change content. At the same time, the line numbers of subordinate projects of "Asset Income Rights" also need to be modified synchronously, including "Loan Income Rights", etc.

[0095] After the data generation script is modified, this embodiment of the invention filters out transaction data belonging to the target project and its subordinate projects from the original transaction data, and generates source data for reporting based on the modified indicator hierarchy in the data generation script. It should be noted that this modification method does not change the calculation logic of the data generation script for the target project and its subordinate projects' transaction data; it only adjusts the format of the source data for reporting.

[0096] Optionally, when the change point is narrowed, the steps for generating the reporting source data include: identifying the project corresponding to the change point as the target project; modifying the calculation logic for the target project in the reporting data generation script based on the name and indicator level of the target project in the change content; filtering out the transaction data of the target project from the original transaction data; and performing calculations on the filtered transaction data based on the modified calculation logic of the reporting data generation script to generate the reporting source data.

[0097] When the change point is reduction, see Table 4 below:

[0098] Table 4

[0099]

[0100] Table 4 shows some of the reports in the data submission specification document. As shown in Table 4, "2024 Narrowing of Statistical Scope" is identified as a change point of the narrowing type. The table header includes: "Line Code", "Item", "RMB", "USD to RMB", "EUR to RMB", "GBP to RMB", "JPY to RMB", "Other Currencies to RMB", "Total", and "Remarks". The changed content includes: "811000", "Loans", and "2024 Narrowing of Statistical Scope".

[0101] When generating the source data for reporting, the project corresponding to the change point is first identified as the target project, i.e., "loan" is identified as the target project. Then, based on the name and indicator level of the target project in the change content, the calculation logic for the target project in the data generation script is modified to... Figure 4 For example, when narrowing down the scope of "loans", "loans" is not the lowest level, but the second level. That is, "loans" is the sum of all items with line number "811***". Therefore, when modifying the data submission specification document, the calculation logic for "loans" in the data submission specification document should be modified to add up the transaction data of all items with line number "811***".

[0102] Then, using the modified data reporting specification document, transaction data belonging to "Loans" and all its subordinate projects are filtered from the original transaction data. Specifically, transaction data for all projects with the transaction number "811***" is filtered. Based on the modified calculation logic of the data reporting generation script, these filtered transaction data are summed to obtain the final source data for the "Loans" project. It should be noted that if there are regulations requiring the removal of certain projects during the data filtering process, such as removing "Repurchase and Interbank Loans" and "Repurchase and Interbank Lending (including Borrowing)" projects, then the transaction data for these two projects need not be included in the data filtering.

[0103] It should be noted that if the target project's indicator level is at the lowest level, other documents are needed to determine which project transaction data should be screened. For example, for "repurchase and interbank lending," the product-by-product data collection specification document is needed to determine which project transaction data should be screened. The product-by-product data collection specification document is a known document issued by regulatory agencies.

[0104] Figure 3 This is a simplified flowchart of an optional source data reporting method based on artificial intelligence according to an embodiment of the present invention, such as... Figure 3As shown, the process begins by obtaining the data submission specification document, using an optical character recognition algorithm to identify text within the document, and then modifying the data submission generation script based on these changes. Next, raw transaction data is retrieved from the underlying data lake of the wealth management subsidiary, and the modified data submission generation script is used to generate the source data for submission. This source data is then format-converted and packaged, and finally sent to the regulatory agency via a dedicated communication line, from which the regulatory agency returns the submission result.

[0105] In this embodiment of the invention, by detecting the reporting data specification file, an optical character recognition algorithm is used to perform text recognition on the reporting data specification file. If target characters or bold characters are found in the recognition results, the area containing the target characters or bold characters is identified as a change point. For each change point, the header and name of the report containing the change point, as well as all information of the row to which the change point belongs, are extracted. The header, name, and all information of the row to which the change point belongs are identified as the change content of the change point. Then, the original transaction data is obtained. Based on the change point and the corresponding change content, the pre-constructed reporting data generation script is modified. Based on the modified reporting data generation script and the original transaction data, reporting source data is generated. Finally, the reporting source data is processed for format conversion and packaging, and the packaged reporting source data is output. This realizes intelligent parsing of the reporting data specification file, automatic location of change points, intelligent modification of the reporting data generation script, and automatic generation and sending of reporting source data.

[0106] The embodiments of the present invention greatly enhance the automation of source data reporting, significantly reduce the investment of R&D resources and manpower, improve the efficiency of source data reporting, and thus solve the technical problem that the source data reporting method in related technologies relies on manual reporting and reporting system modification, resulting in low reporting efficiency.

[0107] The following is a detailed description with reference to another embodiment.

[0108] Example 2

[0109] The source data reporting device based on artificial intelligence provided in this embodiment includes multiple implementation units, each of which corresponds to a specific implementation step in Embodiment 1 above.

[0110] Figure 4 This is a schematic diagram of an optional artificial intelligence-based source data reporting device according to an embodiment of the present invention, such as... Figure 4 As shown, the AI-based source data reporting device may include: an identification unit 41, a first determination unit 42, a second determination unit 43, a generation unit 44, and an output unit 45.

[0111] The recognition unit 41 is used to detect the data submission specification document and uses an optical character recognition algorithm to perform text recognition on the data submission specification document. The content of the data submission specification document is presented in the form of a report.

[0112] The first determining unit 42 is used to determine the area of ​​the target character or bold character as a change point when the target character or bold character exists in the recognition result. The change points include: renaming, adding, adjusting and shrinking.

[0113] The second determining unit 43 extracts the header and name of the report where the change point is located, as well as all information of the row to which the change point belongs, for each change point, and determines the header, name, and all information of the row to which the change point belongs as the change content of that change point.

[0114] The generation unit 44 is used to acquire the original transaction data, modify the pre-built reporting data generation script based on the change points and the corresponding change content, and generate the reporting source data based on the modified reporting data generation script and the original transaction data.

[0115] Output unit 45 is used to perform format conversion and packaging processing on the reported source data, and output the packaged reported source data.

[0116] The aforementioned AI-based source data reporting device detects the reporting data specification document through the recognition unit 41, performs text recognition on the document using an optical character recognition algorithm, and identifies the area containing the target character or bold character as a change point if the recognition result is found by the first determining unit 42. For each change point, the second determining unit 43 extracts the header and name of the report containing the change point, as well as all information of the row to which the change point belongs, and determines the change content of that change point based on the header, name, and row information. Then, the generation unit 44 obtains the original transaction data and, based on the change point and the corresponding change content, modifies the pre-defined data. The constructed reporting data generation script is modified, and based on the modified reporting data generation script and the original transaction data, reporting source data is generated. Finally, the reporting source data is processed by the output unit 45 for format conversion and packaging, and the packaged reporting source data is output. This realizes intelligent parsing of reporting data specification files, automatic location of change points, intelligent modification of reporting data generation script, and automatic generation and sending of reporting source data. This greatly enhances the automation level of source data reporting and improves the efficiency of source data reporting. It also solves the technical problem of low reporting efficiency caused by the reliance on manual reporting and reporting system modification in related technologies.

[0117] Optionally, the recognition unit 41 includes: a first processing module for locating the reported data specification file using a selective search algorithm to generate multiple recognition candidate regions; a second processing module for inputting the reported data specification file into a pre-constructed deep network to obtain a reported feature map; a third processing module for projecting each recognition candidate region onto the reported feature map to generate multiple feature matrices, and inputting each feature matrix into a pooling layer to obtain multiple feature matrices of uniform size; and a fourth processing module for inputting each feature matrix into a fully connected layer and an activation layer to obtain multiple recognition characters, wherein the recognition characters include: target characters and bold characters.

[0118] Optionally, the second determining unit 43 includes: a first determining module, used to determine the change points of the target character and the corresponding change content as the target character change set, and to determine the change points of the bold character and the corresponding change content as the bold character change set; and a comparison module, used to compare the target character change set with the bold character change set, and take the union as the target change set, wherein the target change set is used to modify the pre-built reporting data generation script.

[0119] Optionally, the generation unit 44 includes: a first modification module, used to determine the project corresponding to the change point as the target project when the change point is a name change, and modify the name of the corresponding project in the reporting data generation script based on the serial number and name of the target project in the change content; and a first generation module, used to filter out the transaction data belonging to the target project from the original transaction data, and generate the reporting source data based on the modified project name in the reporting data generation script.

[0120] Optionally, the generation unit 44 further includes: a second modification module, used to determine the project corresponding to the change point as the target project when the change point is newly added, and modify the statistical range of the corresponding project in the reporting data generation script based on the name and statistical range description information of the target project in the change content; and a second generation module, used to filter out transaction data that belongs to the modified statistical range from the original transaction data, and perform calculation processing on the filtered transaction data based on the calculation logic specified in the reporting data generation script to generate reporting source data.

[0121] Optionally, the generation unit 44 further includes: a second determining module, used to determine the project corresponding to the change point as the target project when the change point is an adjustment, and to determine all subordinate projects of the target project; a third modifying module, used to modify the indicator level of the corresponding project in the reporting data generation script based on the name and indicator level of the target project in the change content, wherein the corresponding project includes: the target project and all subordinate projects, and the indicator level is used to describe the classification level of the target project in the reporting system; and a third generating module, used to filter out the transaction data belonging to the target project and subordinate projects from the original transaction data, and generate the reporting source data based on the modified indicator level in the reporting data generation script.

[0122] Optionally, the generation unit 44 further includes: a fourth modification module, used to determine the project corresponding to the change point as the target project when the change point is narrowed, and modify the calculation logic of the target project in the reporting data generation script based on the name and indicator level of the target project in the change content; and a fourth generation module, used to filter out the transaction data of the target project from the original transaction data, and perform calculation processing on the filtered transaction data based on the modified calculation logic of the reporting data generation script to generate the reporting source data.

[0123] The aforementioned artificial intelligence-based source data reporting device may also include a processor and a memory. The aforementioned identification unit 41, first determination unit 42, second determination unit 43, generation unit 44, output unit 45, etc., are all stored in the memory as program units, and the processor executes the aforementioned program units stored in the memory to realize the corresponding functions.

[0124] The aforementioned processor contains a kernel, which retrieves the corresponding program units from memory. One or more kernels can be configured, and source data reporting can be performed based on artificial intelligence by adjusting kernel parameters.

[0125] The aforementioned memory may include non-permanent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0126] Example 3

[0127] Embodiments of this application may provide an electronic device. Figure 5 This is a structural block diagram of an electronic device according to an embodiment of this application. Figure 5 As shown, the electronic device may include: one or more ( Figure 5 Only one of the components is shown: processor 502, memory 504, memory controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module, and display.

[0128] The memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the methods and apparatus in the embodiments of this application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, thereby implementing the above-described methods. The memory may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include memory remotely located relative to the processor, and these remote memories can be connected to the terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0129] The processor can access information and applications stored in memory via a transmission device to execute the following steps: Detecting the reporting data specification file; performing text recognition on the reporting data specification file using an optical character recognition algorithm, wherein the content of the reporting data specification file is presented in report form; if target characters or bold characters are found in the recognition results, the area containing the target characters or bold characters is identified as a change point, including renaming, adding, adjusting, and shrinking; for each change point, extracting the header and name of the report containing the change point, as well as all information of the row to which the change point belongs, and identifying the header, name, and all information of the row to which the change point belongs as the change content of that change point; acquiring the original transaction data; modifying the pre-built reporting data generation script based on the change points and corresponding change content; and generating the reporting source data based on the modified reporting data generation script and the original transaction data; performing format conversion and packaging processing on the reporting source data, and outputting the packaged reporting source data.

[0130] The processor can also access information and applications stored in the memory via a transmission device to perform the following steps: using a selective search algorithm to locate the reported data specification file and generate multiple candidate regions for recognition; inputting the reported data specification file into a pre-built deep network to obtain a reported feature map; projecting each candidate region onto the reported feature map to generate multiple feature matrices, and inputting each feature matrix into a pooling layer to obtain multiple feature matrices of uniform size; inputting each feature matrix into a fully connected layer and an activation layer to obtain multiple recognized characters, wherein the recognized characters include: target characters and bold characters.

[0131] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: determine the change points of the target character and the corresponding change content as the target character change set, and determine the change points of the bold character and the corresponding change content as the bold character change set; compare the target character change set with the bold character change set, and take the union as the target change set, wherein the target change set is used to modify the pre-built reporting data generation script.

[0132] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: determine the project corresponding to the change point as the target project, modify the name of the corresponding project in the reporting data generation script based on the serial number and name of the target project in the change content; filter out the transaction data belonging to the target project from the original transaction data, and generate the reporting source data based on the modified project name in the reporting data generation script.

[0133] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: determine the item corresponding to the change point as the target item; modify the statistical range of the corresponding item in the reporting data generation script based on the name and statistical range description information of the target item in the change content; filter out the transaction data that belongs to the modified statistical range from the original transaction data; and perform calculation processing on the filtered transaction data based on the calculation logic specified in the reporting data generation script to generate the reporting source data.

[0134] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: determine the project corresponding to the change point as the target project, and determine all subordinate projects of the target project; based on the name and indicator level of the target project in the change content, modify the indicator level of the corresponding project in the reporting data generation script, wherein the corresponding project includes: the target project and all subordinate projects, and the indicator level is used to describe the classification level of the target project in the reporting system; filter out the transaction data belonging to the target project and subordinate projects from the original transaction data, and generate the reporting source data based on the modified indicator level in the reporting data generation script.

[0135] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: determine the project corresponding to the change point as the target project, modify the calculation logic of the target project in the reporting data generation script based on the name and indicator level of the target project in the change content; filter the transaction data of the target project from the original transaction data, and perform calculation processing on the filtered transaction data based on the modified calculation logic of the reporting data generation script to generate the reporting source data.

[0136] This application provides a solution for an artificial intelligence-based source data reporting method. By detecting the reporting data specification document, an optical character recognition (OCR) algorithm is used to perform text recognition on the document. If target characters or bold characters are found in the recognition results, the area containing these characters is identified as a change point. For each change point, the header, name, and all information of the row containing the change point are extracted. The header, name, and row information are then identified as the change content for that change point. Original transaction data is then obtained. Based on the change point and its corresponding change content, a pre-built reporting data generation script is modified. Based on the modified script and the original transaction data, reporting source data is generated. Finally, the source data undergoes format conversion and packaging, and the packaged data is output. This achieves intelligent parsing of the reporting data specification document, automatic location of change points, intelligent modification of the reporting data generation script, and automatic generation and transmission of the source data. This solves the technical problem of low reporting efficiency caused by reliance on manual reporting and system modifications in related technologies.

[0137] Those skilled in the art will understand that Figure 5 The structure shown is for illustrative purposes only. Electronic devices can also be smartphones, tablets, handheld computers, mobile internet devices (MIDs), PADs, and other terminal devices. Figure 5 This does not limit the structure of the aforementioned electronic device. For example, electronic devices may also include components that are more... Figure 5 The more or fewer components shown (such as network interfaces, display devices, etc.), or having the same Figure 5 The different configurations shown.

[0138] Those skilled in the art will understand that all or part of the steps in the various artificial intelligence-based source data reporting methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0139] Example 4

[0140] Embodiments of this application also provide a storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the artificial intelligence-based source data reporting method provided in Embodiment 1.

[0141] According to another aspect of the present invention, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored computer program, wherein, when the computer program is running, it controls the device where the computer-readable storage medium is located to execute any one of the artificial intelligence-based source data reporting methods in Embodiment 1 above.

[0142] Optionally, in this embodiment, the storage medium may be located in any computer terminal in a group of computer terminals in a computer network, or in any mobile terminal in a group of mobile terminals.

[0143] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the source data reporting method based on artificial intelligence described in various embodiments of this application.

[0144] This application also provides a computer program product, including a non-volatile computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the artificial intelligence-based source data reporting method described in various embodiments of this application.

[0145] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0146] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0147] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0148] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0149] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0150] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0151] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A source data reporting method based on artificial intelligence, characterized in that, include: The data submission specification document is inspected, and an optical character recognition algorithm is used to perform text recognition on the data submission specification document. The content of the data submission specification document is presented in the form of a report. If the recognition result contains a target character or a bold character, the area of ​​the target character or the bold character is identified as a change point, wherein the change point includes: renaming, adding, adjusting and shrinking; For each change point, extract the table header and table name of the report where the change point is located, as well as all information of the row to which the change point belongs, and determine the table header, table name, and all information of the row to which the change point belongs as the change content of that change point; Obtain the original transaction data, modify the pre-built reporting data generation script based on the change points and corresponding change content, and generate reporting source data based on the modified reporting data generation script and the original transaction data; The reported source data is processed for format conversion and packaging, and the packaged reported source data is output.

2. The source data reporting method according to claim 1, characterized in that, The steps of performing text recognition on the submitted data specification document using the optical character recognition algorithm include: A selective search algorithm is used to locate the reported data specification file and generate multiple candidate regions for identification. The reported data specification file is input into a pre-constructed deep network to obtain a reported feature map; Each of the identified candidate regions is projected onto the reported feature map to generate multiple feature matrices, and each feature matrix is ​​input into a pooling layer to obtain multiple feature matrices of uniform size; Each of the feature matrices is input into a fully connected layer and an activation layer to obtain multiple recognition characters, wherein the recognition characters include the target character and the bold character.

3. The source data reporting method according to claim 1, characterized in that, After determining the change point and corresponding change content of the target character or the bold character, the process includes: The change points and corresponding change content of the target character are determined as the target character change set, and the change points and corresponding change content of the bold character are determined as the bold character change set; The target character change set is compared with the bold character change set, and the union is taken as the target change set. The target change set is used to modify the pre-built reporting data generation script.

4. The source data reporting method according to claim 1, characterized in that, When the change point is a name change, the step of generating the reporting source data includes: The project corresponding to the change point is identified as the target project. Based on the serial number and name of the target project in the change content, the name of the corresponding project in the data reporting generation script is modified. Transaction data belonging to the target project is filtered out from the original transaction data, and the reporting source data is generated based on the modified project name in the reporting data generation script.

5. The source data reporting method according to claim 1, characterized in that, When the change point is newly added, the step of generating the reporting source data includes: The project corresponding to the change point is identified as the target project. Based on the name and statistical range description information of the target project in the change content, the statistical range of the corresponding project in the data reporting generation script is modified. Transaction data that falls within the modified statistical range is selected from the original transaction data. Based on the calculation logic specified in the reporting data generation script, the selected transaction data is processed to generate the reporting source data.

6. The source data reporting method according to claim 1, characterized in that, When the change point is an adjustment, the step of generating the reported source data includes: The project corresponding to the change point is identified as the target project, and all subordinate projects of the target project are identified. Based on the name and indicator level of the target project in the changed content, modify the indicator level of the corresponding project in the reporting data generation script. The corresponding project includes the target project and all the subordinate projects. The indicator level is used to describe the classification level of the target project in the reporting system. Transaction data belonging to the target project and the subordinate project are filtered out from the original transaction data, and the reporting source data is generated based on the modified indicator hierarchy in the reporting data generation script.

7. The source data reporting method according to claim 1, characterized in that, When the change point is reduced, the step of generating the reporting source data includes: The project corresponding to the change point is identified as the target project. Based on the name and indicator level of the target project in the change content, the calculation logic for the target project in the data reporting generation script is modified. The transaction data of the target project is filtered out from the original transaction data. Based on the modified calculation logic of the reporting data generation script, the filtered transaction data is processed to generate the reporting source data.

8. A source data reporting device based on artificial intelligence, characterized in that, include: The recognition unit is used to detect the reporting data specification file and uses an optical character recognition algorithm to perform text recognition on the reporting data specification file, wherein the content of the reporting data specification file is presented in the form of a report; The first determining unit is used to determine the area of ​​the target character or the bold character as a change point when the recognition result contains a target character or a bold character, wherein the change point includes: renaming, adding, adjusting and shrinking; The second determining unit extracts the header and name of the report where the change point is located, as well as all information of the row to which the change point belongs, for each change point, and determines the header, name, and all information of the row to which the change point belongs as the change content of the change point. The generation unit is used to acquire the original transaction data, modify the pre-built reporting data generation script based on the change point and the corresponding change content, and generate the reporting source data based on the modified reporting data generation script and the original transaction data. The output unit is used to perform format conversion and packaging processing on the reported source data, and output the packaged reported source data.

9. An electronic device, characterized in that, It includes one or more processors and a memory, the memory being used to store one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the artificial intelligence-based source data reporting method according to any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the source data reporting method based on artificial intelligence as described in any one of claims 1 to 7.

Images