Business form generation method and device and computer program product

Abstract

The invention discloses a business form generation method and device and a computer program product. The method relates to the field of artificial intelligence, and comprises the following steps: receiving a to-be-filled business form, and identifying position information of N to-be-filled areas of the business form through a form line detection method; determining a filling rule of each to-be-filled area through the content in the business form; first prompt information is sent out in real time, and the first prompt information comprises position information of the N to-be-filled areas; when a filling operation of the user on the target to-be-filled area is detected, second prompt information is sent out, and the second prompt information comprises service analysis content of the target to-be-filled area, and / or an error prompt that the filling content of the user does not conform to the filling rule and a modification suggestion; and when it is detected that the N to-be-filled areas are completely filled, generating a filled target business form. Through the method and the device, the problem of low business form generation efficiency in related technologies is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of artificial intelligence, in particular, to a business form generation method and device and computer program product. BACKGROUND

[0002] In related technologies, the processing and filling of reports rely on data entry, manual pre-check and basic automation tools. Although these methods simplify the process to some extent, there are still significant limitations when facing complex, diversified and highly accurate reports. The data entry method and automation tools in related technologies mainly rely on pre-set rules or fixed question and answer libraries. When users encounter unfamiliar business options during the filling process, it is difficult to obtain immediate and targeted guidance, increasing the likelihood of filling errors and omissions. Although some report filling systems provide basic voice interaction functions, these functions are often insufficient to fully support visually impaired individuals to independently complete report filling. Visually impaired users have understanding barriers during the filling process, reducing the accuracy and efficiency of filling. Reports contain a large amount of text and numerical information, with various formats and complex content. Processing such information is inefficient and difficult to adapt to rapidly changing business needs, resulting in low pre-check accuracy, a lengthy data processing process, and affecting the efficiency and accuracy of report filling.

[0003] In view of the low efficiency of business form generation in related technologies, there is currently no effective solution. SUMMARY

[0004] The main purpose of the present application is to provide a business form generation method, device and computer program product to solve the problem of low efficiency of business form generation in related technologies.

[0005] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a business form generation method is provided. The method comprises: receiving a business form to be filled, identifying the position information of N to-be-filled areas of the business form by a table line detection method, wherein N is a positive integer; determining the filling rule of each to-be-filled area by the content in the business form, wherein the filling rule is a business rule that the content of the to-be-filled area needs to comply with; issuing a first prompt information in real time, wherein the first prompt information contains the position information of the N to-be-filled areas; when detecting a filling operation of a target to-be-filled area by a user, issuing a second prompt information, wherein the second prompt information contains the business analysis content of the target to-be-filled area, and / or an error prompt and modification suggestion that the filling content of the user does not comply with the filling rule; and generating a filled target business form when detecting that all N to-be-filled areas are filled.

[0006] Optionally, the position information of the N to-be-filled areas of the business form identified by the table line detection method comprises: pre-processing an initial image of the to-be-filled business form to obtain a target image, wherein the pre-processing comprises at least one of the following: distortion correction, light equalization, and noise filtering; inputting the target image into an image segmentation model to obtain a heat map of the business form, wherein the image segmentation model comprises an encoder and a decoder, the encoder is configured to capture features of the target image, and the decoder is configured to generate the heat map based on the features; processing the heat map by a preset image processing method to obtain a processed heat map, and performing topological analysis on the processed heat map to obtain the position information of the N to-be-filled areas, wherein the preset image processing method comprises at least one of the following: binary processing, line body repair processing, and dashed line completion processing.

[0007] Optionally, the filling rule of each to-be-filled area is determined based on the content in the business form, which comprises: extracting M fields in the business form by a preset identification method, and constructing a field association graph based on the M fields, wherein the preset identification method comprises at least one of the following: character recognition, seal recognition, and two-dimensional code decoding, M is a positive integer, each node in the field association graph represents a field, and each edge is used to represent an association relationship between the connected nodes; determining the filling rule of each to-be-filled area based on the field association graph.

[0008] Optionally, the filling rule of each to-be-filled area is determined based on the field association graph, which comprises: determining a business type to which the business form belongs based on the content in the business form; determining a preset rule corresponding to the business type, wherein the preset rule comprises at least one of the following: a field format rule, a data type rule, a data fluctuation threshold range, and a matching rule between filled data and units; extracting an association relationship between fields from the field association graph, and determining a dependency rule between the to-be-filled area and the associated fields based on the association relationship, wherein the association relationship comprises at least one of the following: a mathematical relationship between data and a time sequence dependency relationship; and determining the preset rule and the dependency rule as the filling rule.

[0009] Optionally, the first prompt information is issued in real time, which comprises: detecting all to-be-filled areas of the business form, and adding a target identifier to a to-be-filled area that has not been filled; adding a preset mark to a to-be-filled area corresponding to a preset field, wherein the preset field comprises at least one of the following: a mandatory field, a data anomaly field, and a verified field.

[0010] Optionally, the second prompt information is sent out when the filling operation of the user on the target to-be-filled region is detected, comprising: identifying the filling content of the user when the filling operation of the user on the target to-be-filled region is detected; determining whether the filling content conforms to the filling rule; and sending out third prompt information when the filling content does not conform to the filling rule, wherein the third prompt information is used to prompt the user of the filling error and to prompt the filling suggestion of the target to-be-filled region.

[0011] Optionally, the identifying the filling content of the user comprises: detecting whether the filling operation of the user has a pause or repeated input; sending out fourth prompt information and displaying the historical filling content or recommended filling content in the target to-be-filled region when the filling operation of the user is detected to have a pause or repeated input, wherein the fourth prompt information is used to prompt the business analysis content of the target to-be-filled region.

[0012] Optionally, the method further comprises: collecting the historical filling records of the business form and the opinion feedback of the user every preset period; determining the optimization strategy of the filling rule based on the historical filling records and the opinion feedback, and updating the filling rule through the optimization strategy to obtain the updated filling rule.

[0013] In order to achieve the above-mentioned purpose, according to another aspect of the present application, a business form generation device is provided. The device comprises: a receiving unit configured to receive a to-be-filled business form, and identify the position information of N to-be-filled regions of the business form through a table line detection method, wherein N is a positive integer; a first determining unit configured to determine the filling rule of each to-be-filled region through the content in the business form, wherein the filling rule is a business rule that needs to be met when filling the content of the to-be-filled region; a first prompting unit configured to send out first prompt information in real time, wherein the first prompt information contains the position information of the N to-be-filled regions; a second prompting unit configured to send out second prompt information when the filling operation of the user on the target to-be-filled region is detected, wherein the second prompt information contains the business analysis content of the target to-be-filled region, and / or the error prompt and modification suggestion when the filling content of the user does not conform to the filling rule; and a generating unit configured to generate the filled target business form when it is detected that the N to-be-filled regions are all filled.

[0014] In the embodiment of the present application, a service form to be filled in is received, the position information of N to-be-filled-in areas of the service form is recognized by a form line detection method, wherein N is a positive integer; the filling rule of each to-be-filled-in area is determined according to the content in the service form, wherein the filling rule is a business rule that needs to be met when filling in the content of the to-be-filled-in area; the first prompt information is sent in real time, wherein the first prompt information contains the position information of the N to-be-filled-in areas; when detecting the filling operation of the user on the target to-be-filled-in area, the second prompt information is sent, wherein the second prompt information contains the business analysis content of the target to-be-filled-in area, and / or the error prompt and modification suggestion that the filling content of the user does not meet the filling rule; in the case that the N to-be-filled-in areas are all filled in, the filled target service form is generated. Through real-time monitoring of the filling behavior of the user, the potential confusion points are actively identified, specific and personalized filling guidance and business rule explanation are provided, potential errors are pre-checked in real time during the filling process, immediate error positioning and correction suggestions are provided through the first prompt information and the second prompt information, and the user modification cost is significantly reduced. The purpose of improving the accuracy and efficiency of the service form filling is achieved, thereby realizing the technical effect of improving the efficiency of the service form generation, and further solving the technical problem of low efficiency of the service form generation. BRIEF DESCRIPTION OF DRAWINGS

[0015] The accompanying drawings, which form a part of the present application, are intended to provide further understanding of the present application, and are used to interpret the illustrative embodiments of the present application and their descriptions, and do not constitute improper limitations to the present application. In the drawings:

[0016] Figure 1 A hardware structure block diagram of a computer terminal (or mobile device) for implementing a service form generation method is shown;

[0017] Figure 2 A flowchart of a service form generation method according to an embodiment of the present application is shown;

[0018] Figure 3 A schematic diagram of a language assistance process according to an embodiment of the present application is shown;

[0019] Figure 4 A schematic diagram of a service form generation system according to an embodiment of the present application is shown;

[0020] Figure 5 A schematic diagram of a service form generation device according to an embodiment of the present application is shown;

[0021] Figure 6 A structural block diagram of an electronic device according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0022] In the following, the technical solutions in the embodiments of the present application will be described clearly and completely in conjunction with the drawings in the embodiments of the present application, and obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort should fall into the scope of protection of the present application.

[0023] It should be noted that the terms "first", "second" and the like in the description and claims of the present application and the above drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in other than the order illustrated or described herein. In addition, the terms "comprise" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device that includes a series of steps or units does not have to be limited to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0024] It should be noted that the collected information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for display, analyzed data, etc.) involved in the present application are information and data authorized by the user or authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of related data comply with relevant laws, regulations and standards, necessary security measures are taken, do not violate public order and good customs, and provide corresponding operation portal for user to choose authorization or refusal. For example, the system and related users or institutions are provided with an interface to provide the user with a corresponding operation portal for the user to choose to agree or refuse the automatic decision result; if the user chooses to refuse, the expert decision process is entered.

[0025] Embodiment 1

[0026] According to the embodiments of the present application, a method for generating a business form is also provided. It should be noted that the steps shown in the flowchart of the drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0027] The method embodiment provided by the embodiment of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1A hardware structure block diagram of a computer terminal (or mobile device) for implementing the business form generation method is shown. As shown in Figure 1 The computer terminal 10 (or mobile device) can include one or more processors 102 (the processor 102 can include, but not limited to, a processing device such as a Microcontroller Unit (MCU) or a Field-Programmable Gate Array (FPGA)), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it can also include a display, an input / output interface (I / O interface), a Universal Serial Bus (USB) port (which can be included as one of the ports in the Business (BUS) port), a network interface, a power supply and / or a camera. Those skilled in the art can understand that Figure 1 The structure shown is only schematic, which does not limit the structure of the above-mentioned electronic device. For example, the computer terminal 10 can include more or fewer components than those shown in Figure 1 or have a different configuration than that shown in Figure 1 .

[0028] It should be noted that the one or more processors 102 and / or other data processing circuits described above can be referred to herein as "data processing circuits" in general. The data processing circuits can be embodied in whole or in part as software, hardware, firmware, or any combination thereof. In addition, the data processing circuits can be a single independent processing module, or any one of the other elements incorporated into the computer terminal 10 (or mobile device) in whole or in part. As referred to in the embodiments of the present application, the data processing circuit is a processor control (for example, selection of a variable resistance terminal path connected to an interface).

[0029] The memory 104 can be used to store software programs and modules of application software, such as program instructions / data storage devices corresponding to the business form generation method in the embodiments of the present application. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, i.e. implements the business form generation method described above. The memory 104 can include a high-speed random access memory, and can also include a non-volatile memory such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 can further include a memory remotely located with respect to the processor 102, which can be connected to the computer terminal 10 through a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0030] 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.

[0031] The display may be, for example, a touchscreen LCD display that allows the user to interact with the user interface of the computer terminal 10 (or mobile device).

[0032] In the above operating environment, this application provides a method for generating business forms. Figure 2 This is a flowchart of a business form generation method provided according to an embodiment of this application, such as... Figure 2 As shown, the method includes:

[0033] Step S201: Receive the business form to be filled in, and identify the location information of the N areas to be filled in the business form using the table line detection method, where N is a positive integer.

[0034] In step S201, the business form can be a financial institution form. This form is a document or table provided by the financial institution using various templates. Customers fill in the required information, including loan applications, credit card applications, etc., to record key information such as the customer's personal information, financial status, and loan application details. These forms include name, address, Social Security number, income information, and debt status. Financial institutions use these forms to assess the customer's credit risk and loan eligibility. These application forms contain important customer information, therefore requiring efficient, accurate, and reliable processing. Simultaneously, to improve customer experience, it is necessary to accelerate the approval process and reduce the risk of human error.

[0035] First, the system receives business forms uploaded by users or generated by the system. Next, it performs image preprocessing and table line detection on the forms. Image preprocessing includes distortion correction, illumination equalization, and noise filtering to improve the accuracy of table recognition. Table line detection is achieved using an improved U-Net network (a deep learning network architecture) to accurately detect table lines, including solid, dashed, or hidden lines. The output of the U-Net network is a heatmap, where each pixel value represents the probability that the location is a table line. The heatmap is binarized using an adaptive thresholding algorithm, converting it into a black and white image to highlight the table lines. For broken or discontinuous lines, a direction-controlled dilation operation is used for repair, enhancing the continuity and clarity of the lines. For dashed lines, a Markov random field model is used for tracking and completion, predicting the location of missing lines by analyzing the state and probability of adjacent pixels. Finally, topological analysis is used to determine the boundary coordinates (x1, y1, x2, y2) of the cells, i.e., the location information of the areas to be filled.

[0036] Step S202: Determine the filling rules for each field to be filled based on the content in the business form. The filling rules are the business rules that the content to be filled in the field must conform to.

[0037] In step S202, various domain entities in the business form, such as amount, date, name, and account information, are identified, which can be achieved through a large language model. The model can understand text and contextual information and distinguish different types of input fields. Next, the mathematical relationships and time series dependencies between entities are extracted. For example, "Total Loan Amount" = "Monthly Payment" × "Loan Period" or "Reporting Period" should be consistent with the date field in the report. Hard validation rules are set, such as data type and range checks, to ensure that the user input is formatted correctly and within a reasonable range. Soft validation rules are set, and by analyzing the fluctuation thresholds of historical data, abnormal values ​​are identified, thereby determining the applicability of the input rules and the rationality of the data.

[0038] The filling rules can be built-in validation rules, covering multiple aspects such as format, logic, and business compliance. These rules can be flexibly invoked and extended according to different types of business forms. After identifying entities and relationships, specific filling rules are determined based on the context and business type of each field to be filled. For example, for the "Income" field, it checks for units (e.g., yuan, ten thousand yuan); for the "Date of Birth" field, it verifies whether it conforms to common date formats (e.g., YYYY-MM-DD). A hot-update mechanism is used, taking effect in real time according to new business rules. Even during the pre-check process, if business rules change, the filling rules are immediately adjusted to comply with the new requirements.

[0039] Step S203: Issue the first prompt message in real time, wherein the first prompt message contains the location information of N areas to be filled in.

[0040] In step S203, intelligent guidance combining virtual and real methods significantly improves the accuracy and efficiency of form filling. For example, virtual prompts (i.e., the first prompt information) are overlaid in real time on real paper or electronic forms, intuitively marking required fields, automatically correcting formatting errors (such as the correct way to write dates and amounts), and using dynamic arrows to quickly locate the filling position. This is suitable for complex forms or scenarios with strict requirements (such as medical prescriptions and legal documents). When logical contradictions are detected (such as age not matching date of birth), abnormal areas are marked and modification suggestions are given. Voice query of filling rules is also supported, helping users reduce filling errors and providing significant assistance to non-professionals and cross-language filling scenarios.

[0041] Step S204: When a user's input operation on the target field to be filled is detected, a second prompt message is issued. The second prompt message includes the business parsing content of the target field to be filled, and / or an error message and modification suggestion that the user's input does not conform to the input rules.

[0042] In step S204, the large language model understands and parses the specific meaning and business context of each field in the business form. Through domain knowledge enhancement, the model can identify various financial terms and business logic, providing users with detailed business analysis to help them understand the form's requirements. Upon detecting a user's action to fill in a target field, a validity check is immediately performed, which may include data type checks, format validation, range restrictions, and logic checks based on business rules. For example, if a user enters a non-numeric character in the "Income" field, the system will immediately prompt and require the correct number. Using a field value probability distribution model built with a Bayesian network, outliers in the input can be identified, such as numerical fluctuations exceeding the 3σ principle of historical data. Based on the detection results, the system will issue a second prompt to the user, including specific errors and suggested corrections.

[0043] When a user's action in the target field is detected, an instantaneous analysis of the business meaning of that field is generated. The requirements and examples are explained to the user via text or voice, ensuring the user fully understands the requirements. If the user's input does not conform to the preset rules, a pop-up window or voice prompt displays specific information about the error. Based on similar cases or preset logic, possible modification suggestions are provided. For example, if the user enters an incorrect format in the "Date of Birth" field, the correct format will be suggested and an example provided. Multilingual support ensures accurate business analysis and error prompts for users speaking different languages. Furthermore, for visually impaired users, a voice query function for entry rules and clear voice error prompts are provided, ensuring all users can complete the entry process without barriers.

[0044] Step S205: If all N fields to be filled are detected to be filled, generate the filled target business form.

[0045] In step S205, the user-filled data in all fields to be filled in the business form is traversed and its validity is verified. This includes checking data type, format, and range, as well as soft logic verification based on historical data fluctuation thresholds (3σ principle). After all the data in all fields to be filled has passed the validity verification, all the filled data is correctly filled into the corresponding fields, and the target business form is generated.

[0046] The business form generation method provided in this application receives a business form to be filled out, identifies the location information of N fields to be filled out using a table line detection method (where N is a positive integer), determines the filling rules for each field based on the content of the business form (the filling rules are the business rules that the content of the field must conform to), and issues a first prompt message in real time, which includes the location information of the N fields to be filled out. When a user's filling operation on a target field is detected, a second prompt message is issued, which includes the business parsing content of the target field and / or an error message indicating that the user's filling content does not conform to the filling rules, along with modification suggestions. When all N fields are filled out, a filled target business form is generated. By monitoring the user's filling behavior in real time, the method proactively identifies potential points of confusion, provides specific and personalized filling guidance and explanations of business rules, and performs real-time pre-detection of potential errors during the filling process. Through the first and second prompt messages, it provides immediate error location and correction suggestions, significantly reducing the user's modification costs. This achieves the goal of improving the accuracy and efficiency of business form filling, thereby improving the technical effect of increasing the efficiency of business form generation and solving the technical problem of low business form generation efficiency.

[0047] To efficiently fill out business forms, the location information of the areas to be filled in the business form is first identified. Optionally, in the business form generation method provided in this application embodiment, identifying the location information of N areas to be filled in the business form by a table line detection method includes: preprocessing the initial image of the business form to be filled in to obtain a target image, wherein the preprocessing includes at least one of the following: distortion correction, illumination equalization, and noise filtering; inputting the target image into an image segmentation model to obtain a heatmap of the business form, wherein the image segmentation model includes an encoder and a decoder, the encoder is used to capture the features of the target image, and the decoder is used to generate a heatmap based on the features; processing the heatmap using a preset image processing method to obtain a processed heatmap; performing topological analysis on the processed heatmap to obtain the location information of N areas to be filled in, wherein the preset image processing method includes at least one of the following: binarization processing, line repair processing, and dashed line completion processing.

[0048] In some embodiments, the image segmentation model can be a U-Net network model. Image segmentation is a task in image processing aimed at dividing an image into different regions or objects. In business form recognition, image segmentation can be used to locate and separate various parts of the form, such as the header, table, signature area, etc. U-Net can efficiently generate segmented images for segmenting various regions in the form. The initial image can be the original scanned / captured image of the business form. Preprocessing can include: ① Distortion correction: eliminating perspective distortion using a thin-plate spline interpolation algorithm; ② Illumination equalization: multi-scale brightness normalization guided by Retinex (retinal image processing method) theory; ③ Noise filtering: adaptive median filter (window size dynamically adjusted from 3-15). Final output: target image (resolution 300 dpi, error ±5 dpi).

[0049] An improved U-Net structure is used: the encoder incorporates a ResNet34 (residual network) skeleton, and the decoder embeds an ECA (Efficient Channel Attention) module. The final U-Net output is a heatmap (line probability 0-1). Pre-defined image processing methods include adaptive thresholding and directionally controllable dilation. The heatmap is binarized using the adaptive thresholding algorithm, followed by line restoration using directionally controllable dilation (3×15 / 15×3 dual-core processing). Finally, dashed lines are completed using a Markov random field model based on line segment tracking, and the location information of the areas to be filled is constructed through topological analysis.

[0050] This embodiment automatically and accurately identifies the location information of N fields to be filled in a business form, providing a solid foundation for subsequent intelligent assisted filling, legality verification, and pre-inspection. It improves the accuracy and efficiency of data entry, optimizes the user experience, brings significant convenience to visually impaired users and users who need to fill in information across languages, significantly reduces the error rate, and improves the overall processing speed and quality of the business process.

[0051] After determining the location information of the areas to be filled, it is also necessary to determine the filling rules for the areas to be filled. Optionally, in the business form generation method provided in this application embodiment, determining the filling rules for each area to be filled based on the content of the business form includes: extracting M fields from the business form using a preset recognition method, and constructing a field association graph based on the M fields. The preset recognition method includes at least one of the following: text recognition, seal recognition, and QR code decoding. M is a positive integer. Each node in the field association graph represents a field, and each edge is used to represent the association relationship between the nodes connected by the edge. The filling rules for each area to be filled are determined based on the field association graph.

[0052] In some embodiments, text recognition can be performed within the OCR (Optical Character Recognition) subsystem, following the steps of text detection → orientation correction → CRNN (Convolutional Recurrent Neural Network) recognition → confidence labeling. This includes introducing online handwriting synthesis data augmentation, using a spatial Transformer (a deep learning architecture component) to correct distorted text, achieving anti-interference design and handwritten character support; employing seal recognition technology, involving color space conversion → morphological segmentation → support vector machine classification, and utilizing color space analysis + contour fractal dimension calculation for recognition; performing QR code channel processing, including QR code recognition library decoding → verification and validation → data standardization; and identifying M fields in the business form.

[0053] Constructing a field association graph based on M fields is a crucial step in determining the filling rules. Each node in the graph represents a field, and the edges between nodes represent the logical associations or dependencies between fields. For example, the "income" field in a form might be related to the "occupation" field because a specific occupation may have a preset income range; the "date of birth" field might be related to the "age" field because age needs to be calculated based on the date of birth.

[0054] This embodiment constructs a field association graph, which can automatically adapt to the structure and content of different forms and intelligently determine the filling rules, eliminating the need for manual writing of specific rule sets and reducing maintenance costs.

[0055] After constructing the field association graph, the filling rules are determined based on the field association graph. Optionally, in the business form generation method provided in this application embodiment, determining the filling rules for each area to be filled based on the field association graph includes: determining the business type to which the business form belongs based on the content in the business form; determining the preset rules corresponding to the business type, wherein the preset rules include at least one of the following: field format rules, data type rules, data fluctuation threshold range, and matching rules between the filled data and the unit; extracting the association relationships between fields from the field association graph, and determining the dependency rules between the area to be filled and the associated fields based on the association relationships, wherein the association relationships include at least one of the following: mathematical relationships between data and time series dependencies; and determining the preset rules and dependency rules as the filling rules.

[0056] In some embodiments, the raw text data and field association graph from the business form are used as input data. First, domain entity identification is performed. For example, special handling is set for amount / interest rate / date in the financial field, and code standardization conversion is set in the medical field. Second, relationship extraction is performed to establish mathematical relationships between fields (such as sum(A1:A5)=B6) and construct time series dependencies (continuity during the reporting period). Finally, the logic is validated by setting hard rules: data type / range checks, and soft rules: historical data fluctuation thresholds.

[0057] First, based on the content of the business form, determine the specific business type to which the form belongs, such as loan application, credit card account opening, or check deposit. After determining the business type, load the preset rules related to that business type. Preset rules can include field format rules, data type rules, data fluctuation threshold ranges, and matching rules between entered data and units. Extract the mathematical relationships and time series dependencies between fields from the field association graph. These relationships reflect the inherent logic of the data in the business form, such as the difference between income and expenditure should equal savings, or that loan amounts should not fluctuate significantly within a certain period. Based on the extracted associations, determine the dependency rules between the area to be filled and other related fields. For example, if "loan amount" is entered, then "repayment period" and "monthly payment" may depend on the entered loan amount, and their entry rules will also be affected by the loan amount information. Finally, combine the preset rules and dependency rules to form a complete entry rule. In this way, when users start filling out the form, the information entered by users is validated in real time according to these rules, reducing entry errors and providing more detailed entry guidance for visually impaired users and other user groups.

[0058] For example, for a loan application form, the fields in the form include "loan amount", "repayment period", "loan type", and "monthly payment". The business type is determined to be "personal consumption loan". Preset rules applicable to personal consumption loans are loaded, such as the "loan amount" should be a positive number and not exceed 1 million, the "loan type" should be selected from a predefined list, the "repayment period" should be between 1 and 36 months, and the "monthly payment" should be calculated based on the "loan amount" and "repayment period" and be a positive value. There is a mathematical relationship between the "loan amount" and the "monthly payment", that is, "monthly payment = loan amount / repayment period", and the "monthly payment" should also take into account factors such as interest; there is a certain constraint relationship between the "repayment period" and the "loan type", and different types of loans may have different maximum repayment periods. Based on the extracted association relationships, it is determined that the filling of the "monthly payment" area depends on the filling of the "loan amount" and the "repayment period", and should conform to a specific calculation formula; the filling of the "repayment period" may be restricted by the selection of the "loan type". The finally formed filling rules will cover all the above-mentioned preset rules and dependency rules, ensuring the logical consistency of the filled data such as loan amount, repayment period, monthly payment, etc., and at the same time complying with the specific format and range requirements of the loan business.

[0059] After determining the filling rules, the business form is sent into the anomaly detection process. For the anomaly detection of its time-series data, Bayesian inference can be used to output the anomaly probability P; decision thresholds: red alert: P > 0.95; yellow warning: 0.8 < P ≤ 0.95; green normal: P ≤ 0.8.

[0060] This embodiment constructs a complete rule system by combining the business type and the internal logical relationships between fields. It helps to improve the accuracy and efficiency of data filling, and provides strong support for complex data pre-check and anomaly detection. Through real-time rule verification and assisted filling, the security and compliance of data are guaranteed.

[0061] To improve the efficiency of users filling out business forms, the position of the areas to be filled is prompted to the user through the first prompt message. Optionally, in the business form generation method provided in this embodiment of the application, the real-time sending of the first prompt message includes: detecting all the areas to be filled in the business form, adding target identifiers to the unfilled areas to be filled; adding preset marks to the areas to be filled corresponding to preset fields, where the preset fields include at least one of the following: required fields, data anomaly fields, and verified fields.

[0062] In some embodiments, the entire business form is first scanned to identify all areas awaiting user input. These areas may include text input boxes, drop-down menus, checkboxes, etc., which are key points where users need to provide data. For unfilled areas, target markers are automatically added. These markers can be visual, such as displaying a red border or flashing effect around the area, or auditory, such as providing voice prompts to the user about which areas are still unfilled. The purpose of the target markers is to draw the user's attention to incomplete areas and prevent omissions. Preset markers are added to specific types of unfilled areas; these markers are more specific than target markers and provide more detailed information. Preset fields include required fields, data exception fields, and validated fields.

[0063] For example, the system detects all unfilled areas in a business form. It adds flashing red boxes around required fields such as "Source of Income" and "Expense Items." When "-10000" (a negative number) is entered in the "Loan Amount" field, the system immediately detects the anomaly and displays a yellow exclamation mark around the field, along with a voice prompt stating "The loan amount should be positive," helping to correct the error promptly. After correctly filling in and confirming the "Annual Income" field, the system immediately verifies whether the data conforms to preset rules (such as being positive, not exceeding a certain range, etc.). If the data passes validation, the validated field will display a green checkmark, indicating that the data was entered correctly.

[0064] In addition, a multi-turn dialogue state machine (with a three-stage misunderstanding exit mechanism) is used to assist people with disabilities in filling out business forms. Figure 3 This is a schematic diagram of a language assistance process provided according to an embodiment of this application, such as... Figure 3 As shown, in the initial state when the business form is not filled out, a greeting is sent to the user. When the user queries a field, for fields with clear descriptions, suggestions for values ​​to fill in are given directly. For fields with vague descriptions, the user is given supplementary information through contextual clarification before suggestions for values ​​to fill in are provided. After accepting the suggestions, the user fills in the values. The user can also choose to ask further questions while querying a field, and the user will be prompted to fill in the values ​​via voice reply. Once the filling is complete, the assisted filling process ends.

[0065] This embodiment effectively improves the accuracy of business form completion and user experience by issuing real-time prompts. Through visual and auditory feedback, users can clearly understand which areas need to be filled in, which entries need correction, and which entered information is correct. This reduces the time users spend filling out forms and lowers the possibility of data errors.

[0066] By verifying the user's input in real time, a second prompt message is sent to the user. Optionally, in the business form generation method provided in this application embodiment, when a user's input operation on the target field to be filled is detected, sending the second prompt message includes: identifying the user's input content when the user's input operation on the target field to be filled is detected; determining whether the input content conforms to the input rules; and issuing a third prompt message when the input content does not conform to the input rules. The third prompt message is used to prompt the user that the input is incorrect and to provide suggestions for filling in the target field to be filled.

[0067] In some embodiments, although some applications have integrated voice interaction functions, these functions are usually quite basic, lacking real-time voice feedback and intelligent prompts, making it difficult to fully support visually impaired individuals in independently completing report filling. Visually impaired users often face difficulties in obtaining information, high error rates, and low operational efficiency during the filling process. Related technologies mainly rely on screen readers, which cannot provide real-time voice interaction and intelligent prompts. The inability to provide dynamic voice feedback based on the user's input makes it difficult for the user to accurately understand the current filling status. Visually impaired users need to rely on assistance from others or perform repeated operations, significantly reducing filling efficiency. This embodiment combines speech recognition and speech synthesis technologies to provide real-time voice interaction for visually impaired users. For example, when a user inputs "source of income" via voice, the system can immediately provide feedback on the recognized content and, when an error is detected, provide modification suggestions through voice prompts (third prompt information). Furthermore, based on the user's filling history, it intelligently recommends frequently used options, reducing filling errors and improving efficiency.

[0068] Traditional pre-detection methods in related technologies check reports after submission, failing to detect potential errors in real-time during the filling process. Furthermore, error messages are often simplistic and lack specific correction suggestions. This delayed pre-detection approach not only increases the user's modification costs but may also result in low overall report quality. Error detection typically occurs after submission, making it impossible to identify and correct errors in real-time during the filling process. The provided error information is often too general, making it difficult for users to quickly find and correct errors based on the prompts. It also fails to provide targeted modification suggestions based on historical data and context, forcing users to troubleshoot errors themselves. This embodiment utilizes a deep learning model for real-time pre-detection during the user's filling process, identifying potential errors and providing specific modification suggestions. For example, when a user is filling in "Expense Items," the model detects a mismatch between the amount and the item category in real-time and suggests adjusting the category via pop-up windows or voice prompts. This real-time pre-detection function not only improves the accuracy of report filling but also significantly reduces the user's modification costs.

[0069] The real-time detection and assisted filling mechanism in this embodiment monitors user filling behavior, instantly identifies the content being filled, and then performs compliance checks according to preset filling rules. If the content is detected to be inconsistent with the rules, the system immediately provides the user with error messages and filling suggestions to help them correct errors, avoid subsequent pre-check errors and submission delays, and improve the accuracy of form filling and user experience.

[0070] By detecting the user's input operations, business analysis content is provided in real time. Optionally, in the business form generation method provided in this application embodiment, identifying the user's input content includes: detecting whether the user's input operation is paused or repeated; when the user's input operation is detected to be paused or repeated, issuing a fourth prompt message and displaying historical input content or recommended input content in the target input area, wherein the fourth prompt message is used to prompt the business analysis content of the target input area.

[0071] In some embodiments, during the report filling process in related technologies, users often encounter situations where they do not understand business options or fill them out incorrectly. Although intelligent customer service systems can provide some assistance, they often rely on preset rules or fixed question-and-answer databases, and cannot provide personalized explanations and guidance based on the user's real-time filling behavior and context. This passive assistance method is not only inefficient, but may also lead to user errors or omissions. It requires users to actively ask questions and lacks the ability to proactively identify user confusion. It cannot provide targeted explanations based on the content currently filled out by the user and the context. The explanations provided are often too brief and cannot meet the user's need for a deeper understanding of complex business options. This embodiment, through deep learning technology that integrates large language models, can monitor the user's filling behavior in real time, proactively identify potential points of confusion, and provide detailed and personalized explanations through natural language generation technology. For example, when a user encounters difficulties filling out "source of income," the system can automatically detect the user's pauses or repeated input and immediately provide specific explanations about different income source categories to help the user fill in accurately.

[0072] Existing report filling systems suffer from numerous shortcomings in user experience. They lack real-time guidance and personalized support during the filling process, leading to errors or omissions and a poor overall experience. The lack of real-time guidance and feedback during the filling process easily causes confusion and errors. The inability to provide customized suggestions based on user filling history and preferences results in a monotonous user experience. Complex interfaces and operation processes increase the difficulty of use and reduce filling efficiency. This embodiment, through intelligent communication and real-time pre-checking functions, provides personalized guidance and feedback, significantly improving the user experience. For example, the system can intelligently recommend frequently used options based on the user's filling history and provide specific modification suggestions through various means (such as pop-ups and voice prompts) when filling errors are detected. Furthermore, the system supports multilingual and multi-format report filling to meet the needs of different users.

[0073] This embodiment significantly improves the user's typing experience and efficiency by monitoring user typing behavior in real time, especially when pauses or repeated input are detected, and intelligently issuing prompts containing business analysis content, and displaying historical or recommended typing content.

[0074] To ensure the timeliness and accuracy of the filling rules, the filling rules need to be updated regularly. Optionally, in the business form generation method provided in this application embodiment, the method further includes: collecting historical filling records of business forms and user feedback at preset intervals; determining an optimization strategy for the filling rules based on the historical filling records and user feedback; and updating the filling rules through the optimization strategy to obtain the updated filling rules.

[0075] In some embodiments, business reports typically involve large amounts of text and numerical information, with diverse formats and complex content. Data validation tools in related technologies are inefficient at handling such complex data and struggle to adapt to rapidly changing business needs, resulting in low accuracy and efficiency in pre-detection. When processing large-scale, diverse data, they often face problems of insufficient computing resources and slow processing speeds. The inability to quickly adapt to new report formats and content changes limits the applicability of the pre-detection function. When processing complex data, false positives or false negatives are prone to occur, affecting the overall quality of the reports. This embodiment utilizes a deep learning model to improve the accuracy and efficiency of pre-detection. Through parallel processing and algorithm optimization, it can efficiently process large-scale, diverse data in a short time. Furthermore, the model can continuously learn new report formats and content changes through machine learning algorithms, maintaining the flexibility and adaptability of the pre-detection function. It continuously learns new report formats and content changes, providing continuously optimized pre-detection services. The model automatically analyzes historical data and user feedback through machine learning algorithms to optimize pre-detection logic and error message strategies. In addition, it supports automatic updates and maintenance, reducing the need for manual intervention and significantly improving the system's adaptability and maintenance efficiency.

[0076] This embodiment continuously optimizes the filling rules by regularly collecting and analyzing historical filling records and user feedback, thereby improving the accuracy and efficiency of the filling process and enhancing the user experience.

[0077] According to another embodiment of this application, a business form generation system is also provided. Figure 4 This is a schematic diagram of a business form generation system provided according to an embodiment of this application, such as... Figure 4 As shown, the system includes: an intelligent perception layer, an intelligent parsing layer, and an interactive processing layer.

[0078] The intelligent perception layer consists of a dynamic table segmentation algorithm, a multimodal data fusion unit, and a context-aware module. It employs an improved U-Net network combined with morphological operations to achieve complex table line detection (including dashed / hidden line recognition), integrating OCR text recognition (supporting handwritten character recognition), seal feature extraction, and QR code decoding. An attention mechanism is used to analyze the table's semantic field and establish an inter-field association graph.

[0079] The intelligent parsing layer consists of a semantic parsing engine, a dynamic validation rule base, and an intelligent inference unit: it constructs a domain-knowledge-enhanced parsing model based on a deep learning model. It includes pre-defined validation rules (format / logic / business compliance) and supports user-defined extensions. It employs a Bayesian network to construct a probability distribution model of field values ​​to achieve outlier detection.

[0080] The interaction processing layer includes: an AR (Augmented Reality) assisted completion interface, a multi-turn dialogue engine, and an intelligent correction suggestion system. It uses augmented reality technology to visualize error localization, supports context-preserving interaction in natural language question answering, and provides correction suggestions based on similar cases.

[0081] In data management during storage, a blockchain-based evidence storage module, intelligent routing and distribution system, and version tracking system are used: key field data is hashed and uploaded to the blockchain. Different business systems are automatically integrated based on table type. Table modification history is recorded and audit logs are generated.

[0082] This embodiment proposes a bank report pre-checking system based on intelligent communication, real-time pre-checking, and visually impaired-friendly design, by integrating deep learning technology into a business form generation system. This system can not only identify user confusion regarding business options in real time and provide detailed explanations, but also offer a user-friendly filling experience for visually impaired users through voice interaction technology. Furthermore, the system pre-checks for potential errors in real time during the filling process and provides specific modification suggestions, significantly improving the accuracy and efficiency of report completion. Through continuous optimization and adaptability, the system can continuously learn new report formats and content changes, maintaining its efficiency and accuracy. These innovative solutions will bring revolutionary changes to bank report filling, significantly improving user experience and report quality.

[0083] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0084] Example 2

[0085] This application also provides a business form generation apparatus. It should be noted that the business form generation apparatus of this application can be used to execute the business form generation method provided in this application. The business form generation apparatus provided in this application is described below.

[0086] According to an embodiment of this application, an apparatus for implementing the above-described business form generation method is also provided. Figure 5 This is a schematic diagram of a business form generation device provided according to an embodiment of this application, such as... Figure 5 As shown, the device includes:

[0087] The receiving unit 501 is used to receive the business form to be filled in, and to identify the position information of the N areas to be filled in the business form through the table line detection method, where N is a positive integer;

[0088] The first determining unit 502 is used to determine the filling rules for each field to be filled based on the content in the business form, wherein the filling rules are the business rules that the content to be filled in the field must conform to;

[0089] The first prompt unit 503 is used to issue a first prompt message in real time, wherein the first prompt message contains the location information of N areas to be filled in;

[0090] The second prompting unit 504 is used to issue a second prompting message when it detects that the user has filled in the target field to be filled. The second prompting message includes the business parsing content of the target field to be filled, and / or an error message and modification suggestion that the user's filling content does not conform to the filling rules.

[0091] The generation unit 505 is used to generate a filled target business form when all N fields to be filled are detected to be filled.

[0092] The business form generation apparatus provided in this application embodiment receives a business form to be filled in through a receiving unit 501, and identifies the location information of N areas to be filled in the business form through a table line detection method, where N is a positive integer; a first determining unit 502 determines the filling rules for each area to be filled in based on the content of the business form, wherein the filling rules are the business rules that the content to be filled in the area must conform to; a first prompting unit 503 issues a first prompting message in real time, wherein the first prompting message contains the location information of the N areas to be filled in; a second prompting unit 504 issues a second prompting message when it detects that the user has filled in the target area, wherein the second prompting message contains the business parsing content of the target area to be filled in, and / or an error message and modification suggestion that the user's content does not conform to the filling rules; and a generating unit 505 generates a filled target business form when it detects that all N areas to be filled in have been filled in. By monitoring the user's filling behavior in real time, it proactively identifies potential points of confusion, provides specific and personalized filling guidance and explanations of business rules, and conducts real-time pre-detection of potential errors during the filling process. Through the first and second prompting messages, it provides immediate error location and correction suggestions, significantly reducing the user's modification costs. This achieved the goal of improving the accuracy and efficiency of business form filling, thereby realizing the technical effect of improving the efficiency of business form generation and solving the technical problem of low business form generation efficiency.

[0093] Optionally, in the business form generation apparatus provided in this application embodiment, the receiving unit 501 includes: a preprocessing module, used to preprocess the initial image of the business form to be filled in to obtain a target image, wherein the preprocessing includes at least one of the following: distortion correction, illumination equalization, and noise filtering; an input module, used to input the target image into an image segmentation model to obtain a heatmap of the business form, wherein the image segmentation model includes an encoder and a decoder, the encoder is used to capture the features of the target image, and the decoder is used to generate a heatmap based on the features; and an analysis module, used to process the heatmap using a preset image processing method to obtain a processed heatmap, and to perform topological analysis on the processed heatmap to obtain the location information of N areas to be filled in, wherein the preset image processing method includes at least one of the following: binarization processing, line repair processing, and dashed line completion processing.

[0094] Optionally, in the business form generation apparatus provided in this application embodiment, the first determining unit 502 includes: a first extraction module, used to extract M fields from the business form through a preset recognition method, and construct a field association graph based on the M fields, wherein the preset recognition method includes at least one of the following: text recognition, seal recognition and QR code decoding, M is a positive integer, each node in the field association graph represents a field, and each edge is used to represent the association relationship between the nodes connected by the edge; the first determining module is used to determine the filling rules for each area to be filled based on the field association graph.

[0095] Optionally, in the business form generation apparatus provided in this application embodiment, the first determining module includes: a first determining submodule, used to determine the business type to which the business form belongs based on the content in the business form; a second determining submodule, used to determine the preset rules corresponding to the business type, wherein the preset rules include at least one of the following: field format rules, data type rules, data fluctuation threshold range, and matching rules between the filled data and the unit; an extraction submodule, used to extract the association relationship between fields from the field association graph, and determine the dependency rule between the area to be filled and the associated fields based on the association relationship, wherein the association relationship includes at least one of the following: mathematical relationship between data and time series dependency relationship; and a third determining submodule, used to determine the preset rules and dependency rules as filling rules.

[0096] Optionally, in the business form generation device provided in this application embodiment, the first prompting unit 503 includes: a first detection module, used to detect all fields to be filled in the business form and add target identifiers to the fields to be filled that are not filled in; and a first adding module, used to add preset markers to the fields to be filled in corresponding to preset fields, wherein the preset fields include at least one of the following: required fields, data exception fields, and validated fields.

[0097] Optionally, in the business form generation device provided in this application embodiment, the second prompting unit 504 includes: an identification module, used to identify the user's input content when a user's input operation on the target field to be filled is detected; a first judgment module, used to judge whether the input content conforms to the input rules; and a first prompting module, used to issue a third prompting message when the input content does not conform to the input rules, wherein the third prompting message is used to prompt the user that the input is incorrect and to provide suggestions for filling in the target field to be filled.

[0098] Optionally, in the business form generation device provided in this application embodiment, the identification module includes: a detection submodule, used to detect whether the user's filling operation is paused or repeatedly entered; and a prompting submodule, used to issue a fourth prompting message when the user's filling operation is detected to be paused or repeatedly entered, and to display historical or recommended filling content in the target filling area, wherein the fourth prompting message is used to prompt the business parsing content of the target filling area.

[0099] Optionally, in the business form generation apparatus provided in this application embodiment, the apparatus further includes: a collection unit, used to collect historical filling records of business forms and user feedback at preset intervals; and an update unit, used to determine an optimization strategy for filling rules based on historical filling records and user feedback, and update the filling rules through the optimization strategy to obtain updated filling rules.

[0100] It should be noted that the receiving unit 501, the first determining unit 502, the first prompting unit 503, the second prompting unit 504, and the generating unit 505 correspond to steps S201 to S205 in Embodiment 1. The five units and the corresponding steps implement the same instances and application scenarios, but are not limited to the content disclosed in Embodiment 1. It should be noted that the above modules or units can be hardware or software components stored in memory (e.g., memory 104) and processed by one or more processors (e.g., processors 102a, 102b, ..., 102n). The above modules or units can also be part of a device and run in the computer terminal 10 provided in Embodiment 1.

[0101] Example 3

[0102] Embodiments of this application may provide an electronic device. Figure 6 This is a structural block diagram of an electronic device according to an embodiment of this application. Figure 6 As shown, the electronic device may include: one or more ( Figure 6 (Only one is shown) processor 602, memory 604, memory controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0103] 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.

[0104] The processor can access information and applications stored in the memory via a transmission device to perform the following steps: receiving a business form to be filled out; identifying the location information of N fields to be filled out in the business form using a table line detection method, where N is a positive integer; determining the filling rules for each field to be filled out based on the content of the business form, where the filling rules are the business rules that the content of the field to be filled out must conform to; issuing a first prompt message in real time, where the first prompt message contains the location information of the N fields to be filled out; issuing a second prompt message when a user's filling operation on a target field to be filled out is detected, where the second prompt message contains the business parsing content of the target field to be filled out, and / or an error message and modification suggestion that the user's filling content does not conform to the filling rules; and generating a filled target business form when all N fields to be filled out are detected.

[0105] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: preprocessing the initial image of the business form to be filled to obtain a target image, wherein the preprocessing includes at least one of the following: distortion correction, illumination equalization, and noise filtering; inputting the target image into an image segmentation model to obtain a heatmap of the business form, wherein the image segmentation model includes an encoder and a decoder, the encoder is used to capture the features of the target image, and the decoder is used to generate a heatmap based on the features; processing the heatmap using a preset image processing method to obtain a processed heatmap; performing topological analysis on the processed heatmap to obtain the location information of N areas to be filled, wherein the preset image processing method includes at least one of the following: binarization processing, line repair processing, and dashed line completion processing.

[0106] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: extract M fields from the business form using a preset recognition method, and construct a field association graph based on the M fields, wherein the preset recognition method includes at least one of the following: text recognition, seal recognition and QR code decoding, M is a positive integer, each node in the field association graph represents a field, and each edge is used to represent the association relationship between the nodes connected by the edge; determine the filling rules for each area to be filled based on the field association graph.

[0107] The processor can also invoke information and applications stored in the memory via a transmission device to perform the following steps: determine the business type of the business form based on its content; determine the preset rules corresponding to the business type, wherein the preset rules include at least one of the following: field format rules, data type rules, data fluctuation threshold range, and matching rules between the filled data and the unit; extract the relationships between fields from the field association graph, and determine the dependency rules between the area to be filled and the associated fields based on the relationships, wherein the relationships include at least one of the following: mathematical relationships between data and time series dependencies; and determine the preset rules and dependency rules as the filling rules.

[0108] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: detect all fields to be filled in the business form, add target identifiers to the fields to be filled in that are not filled in; add preset markers to the fields to be filled in that are corresponding to preset fields, wherein the preset fields include at least one of the following: required fields, data exception fields, and validated fields.

[0109] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: when a user's filling operation on the target area to be filled is detected, the user's filling content is identified; it is determined whether the filling content conforms to the filling rules; if the filling content does not conform to the filling rules, a third prompt message is issued, wherein the third prompt message is used to prompt the user that the filling is incorrect and to provide suggestions for filling the target area to be filled.

[0110] The processor can also call the information and application stored in the memory through the transmission device to perform the following steps: detect whether there is a pause or repeated input in the user's filling operation; if a pause or repeated input is detected in the user's filling operation, issue a fourth prompt message and display the historical or recommended filling content in the target filling area, wherein the fourth prompt message is used to prompt the business parsing content of the target filling area.

[0111] The processor can also access information and applications stored in the memory via a transmission device to perform the following steps: collect historical data of business forms and user feedback at preset intervals; determine optimization strategies for the filling rules based on the historical data and user feedback, and update the filling rules through the optimization strategies to obtain the updated filling rules.

[0112] This application provides a method for receiving a business form to be filled out, identifying the location information of N fields to be filled out using a table line detection method (where N is a positive integer), determining the filling rules for each field based on the content of the business form (the filling rules are the business rules that the content of the field must conform to), issuing a first prompt in real time (the first prompt includes the location information of the N fields), issuing a second prompt when a user fills out a target field (the second prompt includes the business parsing content of the target field, and / or an error message indicating that the user's content does not conform to the filling rules, along with modification suggestions), and generating a filled target business form when all N fields are filled out. By monitoring the user's filling behavior in real time, proactively identifying potential points of confusion, providing specific and personalized filling guidance and explanations of business rules, and pre-detecting potential errors during the filling process, this method provides immediate error location and correction suggestions through the first and second prompts, significantly reducing the user's modification costs. This achieves the goal of improving the accuracy and efficiency of business form filling, thereby improving the technical effect of increasing business form generation efficiency and solving the technical problem of low business form generation efficiency.

[0113] Those skilled in the art will understand that Figure 6 The structure shown is for illustrative purposes only. Electronic devices can also be smartphones, tablets, handheld computers, mobile internet devices (MIDs), PADs, and other terminal devices. Figure 6 This does not limit the structure of the aforementioned electronic device. For example, electronic devices may also include components that are more... Figure 6 The more or fewer components shown (such as network interfaces, display devices, etc.), or having the same Figure 6 The different configurations shown.

[0114] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0115] Example 4

[0116] 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 business form generation method provided in Embodiment 1.

[0117] 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.

[0118] This application also provides a computer program product that, when executed on a data processing device, is suitable for performing the steps of a business form generation method.

[0119] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0120] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0121] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0122] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0123] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0124] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0125] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for generating business forms, characterized in that, include: Receive a business form to be filled out, and identify the location information of N fields to be filled out in the business form using a table line detection method, where N is a positive integer; The filling rules for each field to be filled are determined based on the content in the business form, wherein the filling rules are the business rules that the content to be filled in the field to be filled must conform to; A first prompt message is issued in real time, wherein the first prompt message contains the location information of the N areas to be filled in; When a user's input operation on the target field is detected, a second prompt message is issued, wherein the second prompt message includes the business parsing content of the target field and / or an error message and modification suggestions that the user's input does not conform to the input rules; Once it is detected that all N fields to be filled have been filled, a filled target business form is generated.

2. The method according to claim 1, characterized in that, The location information of the N fields to be filled in the business form is identified by the table line detection method, including: The initial image of the business form to be filled in is preprocessed to obtain the target image, wherein the preprocessing includes at least one of the following: distortion correction, illumination equalization and noise filtering; The target image is input into an image segmentation model to obtain a heatmap of the business form. The image segmentation model includes an encoder and a decoder. The encoder is used to capture the features of the target image, and the decoder is used to generate the heatmap based on the features. The heatmap is processed by a preset image processing method to obtain a processed heatmap. Topological analysis is then performed on the processed heatmap to obtain the location information of the N regions to be filled. The preset image processing method includes at least one of the following: binarization processing, line repair processing, and dashed line completion processing.

3. The method according to claim 1, characterized in that, The rules for filling in each field to be filled are determined based on the content of the business form, including: M fields are extracted from the business form using a preset recognition method, and a field association graph is constructed based on the M fields. The preset recognition method includes at least one of the following: text recognition, seal recognition, and QR code decoding. M is a positive integer. Each node in the field association graph represents a field, and each edge is used to represent the association relationship between the nodes connected by the edge. The filling rules for each area to be filled are determined based on the field association graph.

4. The method according to claim 3, characterized in that, The filling rules for each region to be filled are determined based on the field association graph, including: The business type to which the business form belongs is determined based on the content in the business form; Determine the preset rules corresponding to the business type, wherein the preset rules include at least one of the following: field format rules, data type rules, data fluctuation threshold range, and matching rules between the entered data and the unit; The relationships between fields are extracted from the field association graph, and the dependency rules between the area to be filled and the associated fields are determined based on the relationships. The relationships include at least one of the following: mathematical relationships between data and time series dependencies. The preset rules and the dependency rules are determined as the filling rules.

5. The method according to claim 1, characterized in that, The first alert message issued in real time includes: Detect all fields to be filled in the business form, and add target identifiers to the fields that are not filled in. Add a preset marker to the area to be filled corresponding to the preset field, wherein the preset field includes at least one of the following: a required field, a data exception field, and a validated field.

6. The method according to claim 1, characterized in that, When a user's action to fill in the target field is detected, a second prompt message is issued, including: Upon detecting a user's input operation on the target field to be filled, the user's input content is identified; Determine whether the entered content conforms to the entered content rules; If the filled content does not conform to the filling rules, a third prompt message is issued, wherein the third prompt message is used to indicate to the user that the filling is incorrect and to provide suggestions for filling in the target area to be filled.

7. The method according to claim 6, characterized in that, Identifying the user's input includes: Detect whether the user's input operation is paused or repeated; If a pause or repeated input is detected in the user's input operation, a fourth prompt message is issued, and historical or recommended input content is displayed in the target input area. The fourth prompt message is used to prompt the business parsing content of the target input area.

8. The method according to claim 1, characterized in that, The method further includes: Collect historical data of completed business forms and user feedback at preset intervals; Based on the historical data and feedback, an optimization strategy for the filling rules is determined, and the filling rules are updated using the optimization strategy to obtain the updated filling rules.

9. A business form generation device, characterized in that, include: The receiving unit is used to receive the business form to be filled in, and to identify the location information of the N areas to be filled in the business form through a table line detection method, where N is a positive integer; The first determining unit is used to determine the filling rules for each of the fields to be filled based on the content in the business form, wherein the filling rules are the business rules that the content of the fields to be filled must conform to; The first prompting unit is used to issue a first prompting message in real time, wherein the first prompting message includes the location information of the N areas to be filled in; The second prompting unit is used to issue a second prompting message when it detects that a user has filled in the target field to be filled. The second prompting message includes the business parsing content of the target field to be filled, and / or an error message and modification suggestion that the user's filling content does not conform to the filling rules. The generation unit is used to generate a filled target business form when it is detected that all of the N fields to be filled have been filled.

10. A computer program product comprising computer instructions, characterized in that, When the computer instructions are executed by the processor, they implement the steps of the business form generation method according to any one of claims 1 to 8.

Images