Business data intelligent identification method and device, equipment, storage medium and product

Abstract

The invention discloses a business data intelligent identification method and device, equipment, a storage medium and a product, and relates to the technical field of data identification, and the method comprises the steps: receiving initial business data uploaded by a user, and carrying out the data processing of the initial business data, and obtaining the processed business data; performing primary intelligent identification on the processed business data through a data identification model to obtain an identification total task corresponding to the processed business data; performing task splitting on the identification general task to obtain identification sub-tasks corresponding to the processed business data; and inputting the identification sub-tasks into a data identification model in parallel, so that the data identification model performs secondary intelligent identification on the processed business data to obtain business element information. According to the method and the device, intelligent identification is performed on the business data after data processing through the data identification model, and meanwhile, the total identification task is split into the identification sub-tasks for parallel identification processing in the identification process, so that the identification efficiency of the business data is improved.

Description

Technical Field

[0001] The present application relates to the field of data identification technology, and more particularly to a method, device, equipment, storage medium and product for intelligent identification of business data. Background Art

[0002] In the current business processing process, users are often required to fill in text or forms according to a preset template format. However, since the templates corresponding to different business types and different business processing institutions are not the same, users need to spend more time and energy to fill in business data when handling different businesses, which makes the efficiency of business data recognition relatively low. Therefore, the industry is in urgent need of a method that can improve the efficiency of business data recognition. Summary of the invention

[0003] The main purpose of this application is to provide a method, device, equipment, storage medium and product for intelligent identification of business data, aiming to solve the technical problem of low efficiency in identifying business data in the prior art.

[0004] To achieve the above purpose, the present application provides a method for intelligently identifying business data, the method comprising the following steps:

[0005] Receiving initial business data uploaded by a user, and performing data processing on the initial business data to obtain processed business data;

[0006] Performing intelligent recognition on the processed business data through a data recognition model to obtain a total recognition task corresponding to the processed business data;

[0007] Splitting the overall recognition task to obtain recognition subtasks corresponding to the processed business data;

[0008] The identification subtasks are input into the data identification model in parallel, so that the data identification model performs secondary intelligent identification on the processed business data to obtain business element information.

[0009] In one embodiment, the step of performing data processing on the initial service data to obtain processed service data includes:

[0010] If the initial business data is text data, performing data cleaning and data correction processing on the text data to obtain processed business data;

[0011] If the initial business data is image data, performing optical character recognition processing, data cleaning processing and data correction processing on the image data to obtain processed business data;

[0012] If the initial business data is document data, text extraction processing, data cleaning processing and data correction processing are performed on the document data to obtain processed business data.

[0013] In one embodiment, the step of performing data processing on the initial service data to obtain processed service data further includes:

[0014] Determining whether the image data and / or the document data contain a preset identification image according to image pixel values ​​in the image data and / or the document data;

[0015] If so, the region containing the preset identification image in the image data and / or the document data is made transparent, and the transparent initial business data is processed to obtain processed business data.

[0016] In one embodiment, the step of performing data processing on the initial service data to obtain processed service data further includes:

[0017] Determining whether the image data and / or the document data contain a preset identification image according to image pixel values ​​in the image data and / or the document data;

[0018] If so, the region containing the preset identification image in the image data and / or the document data is made transparent, and the transparent initial business data is processed to obtain processed business data.

[0019] In one embodiment, before the step of receiving the initial service data uploaded by the user and performing data processing on the initial service data to obtain the processed service data, the step further includes:

[0020] When receiving a service request from a user, performing distributed identity mapping on the identity information of the user to obtain an initial identity identifier corresponding to the user;

[0021] Obfuscating the initial identity to obtain a fuzzy identity, and performing security authentication on the fuzzy identity through a preset form;

[0022] If the fuzzy identity passes the security authentication, receiving the initial business data uploaded by the user;

[0023] If the fuzzy identity fails the security authentication, the business processing connection with the user is disconnected and an alarm is sent to the administrator.

[0024] In addition, to achieve the above purpose, the present application also proposes a data recognition model construction method, the data recognition model construction method comprising the following steps:

[0025] Constructing a sample data set according to the historical business data, the historical recognition tasks corresponding to the historical business data, and the historical business element information corresponding to the historical business data;

[0026] Performing data preprocessing and data splitting on the sample data set to obtain a model training set, a model test set, and a model verification set;

[0027] The initial large language model is trained and optimized based on the model training set, the model test set and the model verification set to obtain a data recognition model.

[0028] In addition, to achieve the above purpose, the present application also proposes a business data intelligent identification device, the business data intelligent identification device comprising:

[0029] A data processing module, used to receive the initial business data uploaded by the user, and perform data processing on the initial business data to obtain processed business data;

[0030] A first recognition module, used to perform intelligent recognition on the processed business data through a data recognition model to obtain a total recognition task corresponding to the processed business data;

[0031] A task splitting module is used to split the overall recognition task to obtain recognition subtasks corresponding to the processed business data;

[0032] The second recognition module is used to input the recognition subtasks into the data recognition model in parallel, so that the data recognition model performs secondary intelligent recognition on the processed business data to obtain business element information.

[0033] In addition, to achieve the above-mentioned purpose, the present application also proposes a business data intelligent identification device, which includes: a memory, a processor, and a business data intelligent identification program stored in the memory and executable on the processor, and the business data intelligent identification program is configured to implement the steps of the business data intelligent identification method described above.

[0034] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a business data intelligent identification program is stored on the storage medium. When the business data intelligent identification program is executed by the processor, the steps of the business data intelligent identification method described above are implemented.

[0035] In addition, to achieve the above objectives, the present invention also provides a computer program product, which includes a business data intelligent identification program, and when the business data intelligent identification program is executed by a processor, it implements the steps of the business data intelligent identification method described above.

[0036] This application receives the initial business data uploaded by the user, and processes the initial business data to obtain the processed business data; performs an intelligent recognition on the processed business data through a data recognition model to obtain the total recognition task corresponding to the processed business data; splits the total recognition task to obtain the recognition subtasks corresponding to the processed business data; and inputs the recognition subtasks into the data recognition model in parallel, so that the data recognition model performs a second intelligent recognition on the processed business data to obtain business element information. Compared with the traditional business data recognition method, the above method of this application performs a first intelligent recognition and a second intelligent recognition on the data processed business data through a data recognition model, and splits the total recognition task into recognition subtasks for parallel recognition processing during the recognition process, thereby improving the recognition efficiency of business data and effectively shortening the business processing cycle. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 A schematic diagram of the structure of a business data intelligent identification device in a hardware operating environment involved in an embodiment of the present application;

[0038] Figure 2 This is a flow chart of the first embodiment of the method for intelligently identifying business data of the present application;

[0039] Figure 3 This is an interactive diagram of the identification process of the intelligent identification method for business data in this application;

[0040] Figure 4 This is a flow chart of the second embodiment of the business data intelligent identification method of the present application;

[0041] Figure 5 This is a flow chart of the third embodiment of the method for intelligently identifying business data of the present application;

[0042] Figure 6 This is a structural block diagram of the first embodiment of the business data intelligent identification device of the present application.

[0043] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0044] It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0045] Reference Figure 1 , Figure 1 This is a schematic diagram of the structure of a business data intelligent identification device in the hardware operating environment involved in the embodiment of the present application.

[0046] like Figure 1 As shown, the business data intelligent identification device may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a wireless fidelity (Wireless-Fidelity, Wi-Fi) interface). The memory 1005 may be a high-speed random access memory (Random Access Memory, RAM), or a stable non-volatile memory (Non-Volatile Memory, NVM), such as a disk storage. The memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0047] Those skilled in the art will understand that Figure 1 The structure shown in the figure does not constitute a limitation on the business data intelligent identification device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0048] like Figure 1 As shown, the memory 1005 as a storage medium may include an operating system, a network communication module, a user interface module, and a business data intelligent identification program.

[0049] exist Figure 1 In the business data intelligent identification device shown, the network interface 1004 is mainly used for data communication with the network server; the user interface 1003 is mainly used for data interaction with the user; the processor 1001 and the memory 1005 in the business data intelligent identification device of the present application can be set in the business data intelligent identification device, and the business data intelligent identification device calls the business data intelligent identification program stored in the memory 1005 through the processor 1001, and executes the business data intelligent identification method provided in the embodiment of the present application.

[0050] The present application provides a method for intelligently identifying business data. Figure 2 , Figure 2 This is a flow chart of the first embodiment of the business data intelligent identification method of the present application.

[0051] In this embodiment, the business data intelligent identification method includes the following steps:

[0052] Step S10: receiving the initial business data uploaded by the user, and performing data processing on the initial business data to obtain processed business data.

[0053] It should be noted that the execution subject of the method of this embodiment can be a terminal device with task splitting, data processing and program running functions, such as a smart phone, a computer, etc., or an electronic device with the same or similar functions, such as the above-mentioned business data intelligent recognition device. The following takes the business data intelligent recognition device (hereinafter referred to as the recognition device) as an example to illustrate this embodiment and the following embodiments.

[0054] It is understandable that the above-mentioned initial business data can be data filled in by the user for the business to be handled, such as financial business data (such as corporate income data), insurance business data (such as risk management data), change business data (such as identity information change data), etc. This embodiment does not limit this.

[0055] In a specific implementation, data processing may be achieved by performing operations such as data cleaning, data correction, data identification, and data extraction on the above-mentioned initial business data, thereby obtaining the above-mentioned processed business data.

[0056] Step S20: Perform intelligent recognition on the processed business data through a data recognition model to obtain a total recognition task corresponding to the processed business data.

[0057] It should be noted that the above data recognition model is a model that can intelligently recognize business data. Specifically, the above data recognition model can be a model obtained by training and optimizing a large language model (LLM).

[0058] It should be understood that the above data recognition model can be used to perform key content retrieval on the processed business data to achieve intelligent recognition, thereby obtaining the overall recognition task corresponding to the processed business data. The overall recognition task represents the task corresponding to the target data that the user needs to obtain in this business handling.

[0059] Step S30: splitting the overall recognition task into tasks to obtain recognition subtasks corresponding to the processed business data.

[0060] It should be noted that in a multi-user, multi-business processing scenario, the task volume of the total identification task is large. Therefore, in this embodiment, the total identification task can be divided into several identification sub-tasks, thereby avoiding data congestion during task processing and improving the identification efficiency of business data.

[0061] In the specific implementation, the above identification task can be split based on the PROMPT method. P stands for Purpose, which is used to clarify the overall purpose of the task; R stands for Result, which is used to determine the results or standards required to complete the task; O stands for Objectives, which is used to list the main goals required to achieve the task; M stands for Methods, which is used to define the methods or means to achieve the goal; P stands for Processes, which is used to programmatically divide the specific steps and processes; T stands for Time, which is used to determine the time frame of each subtask.

[0062] Step S40: Input the identification subtasks into the data identification model in parallel, so that the data identification model performs secondary intelligent identification on the processed business data to obtain business element information.

[0063] It should be noted that the above-mentioned business element information refers to the necessary information that the user needs to provide when handling a business. For example, assuming that the user needs to handle a transfer business, the business element information obtained after the above-mentioned data recognition model performs a secondary intelligent recognition on the processed business data may include the transfer target account, transfer amount, transfer time, etc.

[0064] In the specific implementation, you can refer to Figure 3 , Figure 3 This is an interactive diagram of the identification process of the business data intelligent identification method of this application. Figure 3 As can be seen in the figure, after the initial business data uploaded by the user is intelligently identified by the data recognition model, the obtained business element information will be fed back to the user so that the user can modify the identified content. In particular, the above business element information can be fed back to the user through the interactive page, and the business element information can be accurately located and highlighted in the initial business data on the interactive page, so that the user can intuitively see the corresponding relationship between the recognition result and the original document, thereby improving the user's efficiency and interactive experience during the operation, as well as the trust in the recognition result.

[0065] This embodiment receives the initial business data uploaded by the user, and processes the initial business data to obtain processed business data; performs an intelligent recognition on the processed business data through a data recognition model to obtain a total recognition task corresponding to the processed business data; splits the total recognition task to obtain recognition subtasks corresponding to the processed business data; and inputs the recognition subtasks into the data recognition model in parallel, so that the data recognition model performs a second intelligent recognition on the processed business data to obtain business element information. Compared with the traditional business data recognition method, the above method of this embodiment performs a first intelligent recognition and a second intelligent recognition on the data processed business data through a data recognition model, and splits the total recognition task into recognition subtasks for parallel recognition processing during the recognition process, thereby improving the recognition efficiency of business data and effectively shortening the business processing cycle.

[0066] refer to Figure 4 , Figure 4 This is a flow chart of the second embodiment of the business data intelligent identification method of the present application.

[0067] In a feasible implementation manner, the step S10 may include:

[0068] Step S101: If the initial business data is text data, data cleaning and data correction are performed on the text data to obtain processed business data.

[0069] It should be understood that the above-mentioned data cleaning process may include removing illegal words, normalizing upper and lower cases, removing duplicate data, etc., and the above-mentioned data correction process may include spelling error correction, grammatical error correction, format error correction, etc., which is not limited in this embodiment.

[0070] Step S102: If the initial business data is image data, optical character recognition processing, data cleaning processing and data correction processing are performed on the image data to obtain processed business data.

[0071] It is understandable that optical character recognition (OCR) refers to a technology for searching, extracting and identifying text in images, determining its shape by detecting dark and light patterns, and then translating the shape into computer text using character recognition methods.

[0072] In a specific implementation, the above-mentioned image data can be subjected to optical character recognition processing based on the following steps. The first step is to analyze the layout in the image data to obtain the text area. The second step is to segment the text area into individual characters for character recognition. The third step is to use a classifier to recognize the segmented characters and convert them into editable text. The fourth step is to post-process the recognition results, such as correcting errors, correcting tilt, removing redundancy, etc., so as to obtain the above-mentioned processed business data.

[0073] Step S103: If the initial business data is document data, text extraction processing, data cleaning processing and data correction processing are performed on the document data to obtain processed business data.

[0074] Step S104: judging whether the image data and / or the document data contain a preset identification image according to the image pixel values ​​in the image data and / or the document data.

[0075] It should be noted that the above-mentioned preset logo image can be a seal logo, a positioning logo, etc., and this embodiment does not limit this.

[0076] Step S105: If yes, the region containing the preset identification image in the image data and / or the document data is made transparent, and the transparent initial business data is processed to obtain processed business data.

[0077] In a specific implementation, taking the preset logo image as a red seal logo as an example, the red seal logo can be located by analyzing the RGB pixel values ​​in the image data and / or document data. This process not only involves accurately identifying the red area, but also replacing or making these pixels transparent so that the originally obscured text becomes clearer, and then processing the transparent initial business data to obtain more comprehensive processed business data.

[0078] In a feasible implementation manner, the step S30 may include:

[0079] Step S301: determine the target output data corresponding to the overall recognition task, and perform logical inversion on the target output data to obtain the target input data corresponding to the overall recognition task.

[0080] Step S302: splitting the overall recognition task based on the target input data to obtain recognition subtasks corresponding to the processed business data.

[0081] In a feasible implementation manner, before step S10, the following may also be included:

[0082] Step S1: When a service request from a user is received, a distributed identity mapping is performed on the identity information of the user to obtain an initial identity identifier corresponding to the user.

[0083] In a specific implementation, a hash algorithm (such as SHA-256) can be used to perform distributed identity mapping on the user's identity information, and the generated hash value can be used as the initial identity identifier corresponding to the user.

[0084] Step S2: Obtain a fuzzy identity after blurring the initial identity, and perform security authentication on the fuzzy identity through a preset form.

[0085] It should be noted that the above-mentioned preset form may be a form containing fuzzy identity identifiers corresponding to several abnormal users. In particular, in order to prevent the preset form from being tampered with, a consortium blockchain may be introduced to store and share the preset form.

[0086] In a specific implementation, the initial identity identifier may be obfuscated based on data hiding, for example, hiding the middle four or more digits of the initial identity identifier and retaining only the front and rear digits; the initial identity identifier may also be obfuscated based on data replacement, for example, replacing some characters with specified characters, etc. This embodiment does not specifically limit the implementation method of obfuscation.

[0087] Step S3: If the fuzzy identity passes the security authentication, the initial business data uploaded by the user is received.

[0088] Step S4: If the fuzzy identity fails the security authentication, the business processing connection with the user is disconnected and an alarm is sent to the administrator.

[0089] In this embodiment, if the initial business data is text data, data cleaning processing and data correction processing are performed on the text data to obtain processed business data; if the initial business data is image data, optical character recognition processing, data cleaning processing and data correction processing are performed on the image data to obtain processed business data; if the initial business data is document data, text extraction processing, data cleaning processing and data correction processing are performed on the document data to obtain processed business data; according to the image pixel values ​​in the image data and / or the document data, it is determined whether the image data and / or the document data contain a preset identification image; if so, the area of ​​the image data and / or the document data containing the preset identification image is made transparent, and data processing is performed on the transparent initial business data. Obtain the processed business data; determine the target output data corresponding to the overall recognition task, and perform logical inversion on the target output data to obtain the target input data corresponding to the overall recognition task; perform task splitting on the overall recognition task based on the target input data to obtain the recognition subtasks corresponding to the processed business data; when receiving a business request from a user, perform distributed identity mapping on the identity information of the user to obtain the initial identity identifier corresponding to the user; after blurring the initial identity identifier, obtain a fuzzy identity identifier, and perform security authentication on the fuzzy identity identifier through a preset form; if the fuzzy identity identifier passes the security authentication, receive the initial business data uploaded by the user; if the fuzzy identity identifier fails the security authentication, disconnect the business processing connection with the user, and send an alarm to the administrator. Compared with the traditional business data recognition method, the above method of this embodiment integrates important technologies such as optical character recognition, large language model, distributed identity mapping, and data blurring, which reduces manual intervention while ensuring the efficiency and security of data recognition, thereby bringing users a smoother and more convenient business handling experience.

[0090] refer to Figure 5 , Figure 5 This is a flow chart of the first embodiment of the method for building a data recognition model of the present application.

[0091] In this embodiment, the data recognition model construction method includes the following steps:

[0092] Step S1 ′: construct a sample data set according to the historical business data, the historical identification tasks corresponding to the historical business data, and the historical business element information corresponding to the historical business data.

[0093] It should be understood that for the data recognition model in this embodiment, when the model input is business data, the model output is the historical recognition task corresponding to the business data; when the model input is business data and the historical recognition task corresponding to the business data, the model output is business element information. Therefore, when constructing a sample data set, it is necessary to collect the above-mentioned historical business data, the historical recognition task corresponding to the historical business data, and the historical business element information corresponding to the historical business data.

[0094] Step S2': perform data preprocessing and data splitting on the sample data set to obtain a model training set, a model test set and a model verification set.

[0095] It should be noted that the above-mentioned model training set is used to train the model, that is, to determine the parameters of the model; the above-mentioned model test set is used to evaluate the generalization ability of the model after the model training is completed, that is, the model's ability to predict unknown data; the above-mentioned model validation set is used to evaluate the performance of the model in order to adjust and optimize the parameters.

[0096] It should be understood that the sample data set after data preprocessing can be split based on a simple random split method, a K-fold cross validation method or a stratified sampling method to obtain a model training set, a model test set and a model validation set.

[0097] Step S3': training and optimizing the initial large language model based on the model training set, the model test set and the model verification set to obtain a data recognition model.

[0098] This embodiment constructs a sample data set based on historical business data, historical recognition tasks corresponding to the historical business data, and historical business element information corresponding to the historical business data; performs data preprocessing and data splitting on the sample data set to obtain a model training set, a model test set, and a model verification set; trains and optimizes the initial large language model based on the model training set, the model test set, and the model verification set to obtain a data recognition model. The above method of this embodiment trains and optimizes the initial large language model based on the model training set, model test set, and model verification set constructed based on historical business data, historical recognition tasks corresponding to the historical business data, and historical business element information corresponding to the historical business data, so that the data recognition model finally obtained can intelligently recognize business data, thereby improving the recognition efficiency and accuracy of business data.

[0099] In addition, an embodiment of the present application further proposes a storage medium, on which a business data intelligent identification program is stored. When the business data intelligent identification program is executed by a processor, the steps of the business data intelligent identification method described above are implemented.

[0100] Reference Figure 6 , Figure 6This is a structural block diagram of the first embodiment of the business data intelligent identification device of the present application.

[0101] like Figure 6 As shown, the business data intelligent identification device proposed in the embodiment of the present application includes:

[0102] The data processing module 601 is used to receive the initial business data uploaded by the user and perform data processing on the initial business data to obtain processed business data;

[0103] A first identification module 602 is used to perform an intelligent identification on the processed business data through a data identification model to obtain an overall identification task corresponding to the processed business data;

[0104] The task splitting module 603 is used to split the overall recognition task to obtain recognition subtasks corresponding to the processed business data;

[0105] The second recognition module 604 is used to input the recognition subtasks into the data recognition model in parallel, so that the data recognition model performs secondary intelligent recognition on the processed business data to obtain business element information.

[0106] This embodiment receives the initial business data uploaded by the user, and processes the initial business data to obtain processed business data; performs an intelligent recognition on the processed business data through a data recognition model to obtain a total recognition task corresponding to the processed business data; splits the total recognition task to obtain recognition subtasks corresponding to the processed business data; and inputs the recognition subtasks into the data recognition model in parallel, so that the data recognition model performs a second intelligent recognition on the processed business data to obtain business element information. Compared with the traditional business data recognition method, the above method of this embodiment performs a first intelligent recognition and a second intelligent recognition on the data processed business data through a data recognition model, and splits the total recognition task into recognition subtasks for parallel recognition processing during the recognition process, thereby improving the recognition efficiency of business data and effectively shortening the business processing cycle.

[0107] Based on the first embodiment of the business data intelligent identification device of the present application, a second embodiment of the business data intelligent identification device of the present application is proposed.

[0108] In this embodiment, the data processing module 601 is also used to, if the initial business data is text data, perform data cleaning processing and data correction processing on the text data to obtain processed business data; if the initial business data is image data, perform optical character recognition processing, data cleaning processing and data correction processing on the image data to obtain processed business data; if the initial business data is document data, perform text extraction processing, data cleaning processing and data correction processing on the document data to obtain processed business data.

[0109] Furthermore, the data processing module 601 is also used to determine whether the image data and / or the document data contain a preset identification image based on image pixel values ​​in the image data and / or the document data; if so, the area of ​​the image data and / or the document data containing the preset identification image is made transparent, and the transparent initial business data is processed to obtain processed business data.

[0110] Furthermore, the task splitting module 603 is also used to determine the target output data corresponding to the overall identification task, and perform logical reverse deduction on the target output data to obtain the target input data corresponding to the overall identification task; based on the target input data, the overall identification task is split to obtain the identification sub-tasks corresponding to the processed business data.

[0111] Furthermore, the data processing module 601 is also used to perform distributed identity mapping on the identity information of the user when receiving a business request from the user, so as to obtain the initial identity identifier corresponding to the user; to obtain a fuzzy identity identifier after blurring the initial identity identifier, and to perform security authentication on the fuzzy identity identifier through a preset form; if the fuzzy identity identifier passes the security authentication, the initial business data uploaded by the user is received; if the fuzzy identity identifier fails the security authentication, the business processing connection with the user is disconnected, and an alarm is sent to the administrator.

[0112] Other embodiments or specific implementation methods of the business data intelligent identification device of the present application can refer to the above-mentioned method embodiments and will not be repeated here.

[0113] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.

[0114] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0115] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as a read-only memory / random access memory, a disk, or an optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0116] The above are only preferred embodiments of the present application, and are not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A method for intelligently identifying business data, characterized in that: The method comprises the following steps: Receiving initial business data uploaded by a user, and performing data processing on the initial business data to obtain processed business data; Performing intelligent recognition on the processed business data through a data recognition model to obtain a total recognition task corresponding to the processed business data; Splitting the overall recognition task to obtain recognition subtasks corresponding to the processed business data; The identification subtasks are input into the data identification model in parallel, so that the data identification model performs secondary intelligent identification on the processed business data to obtain business element information.

2. The business data intelligent identification method according to claim 1, characterized in that: The step of processing the initial business data to obtain processed business data includes: If the initial business data is text data, performing data cleaning and data correction processing on the text data to obtain processed business data; If the initial business data is image data, performing optical character recognition processing, data cleaning processing and data correction processing on the image data to obtain processed business data; If the initial business data is document data, text extraction processing, data cleaning processing and data correction processing are performed on the document data to obtain processed business data.

3. The business data intelligent identification method according to claim 2, characterized in that: The step of processing the initial business data to obtain processed business data also includes: Determining whether the image data and / or the document data contain a preset identification image according to image pixel values ​​in the image data and / or the document data; If so, the region containing the preset identification image in the image data and / or the document data is made transparent, and the transparent initial business data is processed to obtain processed business data.

4. The business data intelligent identification method according to claim 1, characterized in that: The step of splitting the overall identification task to obtain identification subtasks corresponding to the processed business data includes: Determine the target output data corresponding to the overall recognition task, and perform logical inversion on the target output data to obtain the target input data corresponding to the overall recognition task; The overall recognition task is split based on the target input data to obtain recognition subtasks corresponding to the processed business data.

5. The method for intelligently identifying business data according to claim 1, characterized in that: Before the step of receiving the initial business data uploaded by the user and performing data processing on the initial business data to obtain the processed business data, the step further includes: When receiving a service request from a user, performing distributed identity mapping on the identity information of the user to obtain an initial identity identifier corresponding to the user; Obfuscating the initial identity to obtain a fuzzy identity, and performing security authentication on the fuzzy identity through a preset form; If the fuzzy identity passes the security authentication, receiving the initial business data uploaded by the user; If the fuzzy identity fails the security authentication, the business processing connection with the user is disconnected and an alarm is sent to the administrator.

6. A method for constructing a data recognition model, characterized in that: The method comprises the following steps: Constructing a sample data set according to the historical business data, the historical recognition tasks corresponding to the historical business data, and the historical business element information corresponding to the historical business data; Performing data preprocessing and data splitting on the sample data set to obtain a model training set, a model test set, and a model verification set; The initial large language model is trained and optimized based on the model training set, the model test set and the model verification set to obtain a data recognition model.

7. A business data intelligent identification device, characterized in that: The business data intelligent identification device comprises: A data processing module, used to receive the initial business data uploaded by the user, and perform data processing on the initial business data to obtain processed business data; A first recognition module, used to perform intelligent recognition on the processed business data through a data recognition model to obtain a total recognition task corresponding to the processed business data; A task splitting module is used to split the overall recognition task to obtain recognition subtasks corresponding to the processed business data; The second recognition module is used to input the recognition subtasks into the data recognition model in parallel, so that the data recognition model performs secondary intelligent recognition on the processed business data to obtain business element information.

8. A business data intelligent identification device, characterized in that: The device comprises: a memory, a processor, and a business data intelligent identification program stored in the memory and executable on the processor, wherein the business data intelligent identification program is configured to implement the steps of the business data intelligent identification method as described in any one of claims 1 to 5.

9. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a business data intelligent identification program is stored on the storage medium. When the business data intelligent identification program is executed by a processor, the steps of the business data intelligent identification method according to any one of claims 1 to 5 are implemented.

10. A computer program product, characterized in that The computer program product comprises a business data intelligent identification program, and when the business data intelligent identification program is executed by a processor, the steps of the business data intelligent identification method according to any one of claims 1 to 5 are implemented.

Images