Form dynamic field generation method and device, electronic equipment and storage medium

Abstract

The invention provides a form dynamic field generation method. The method comprises the following steps: acquiring a form type specified by a user; inputting the form type specified by the user into a trained machine learning model, so that the machine learning model performs association processing on the form type specified by the user and a field, and a field corresponding to the form type specified by the user is obtained; determining whether to trigger a dynamic field generation mechanism or not according to the content filled in the field corresponding to the form type specified by the user by the user; and if the dynamic field generation mechanism is triggered, inputting the fields and the filling contents filled by the user into a trained machine learning model, so that the machine learning model performs association processing on the filled fields, the filling contents and the dynamic fields to obtain predicted dynamic fields. According to the form filling method and device, the targeted dynamic fields can be generated according to the fields selected to be filled in the form by the user and the specific filling content, and the form filling efficiency and adaptability are improved.

Description

Technical Field

[0001] One or more embodiments of the present specification relate to the field of form generation technology, and in particular, to a method, device, electronic device, and storage medium for generating dynamic fields of a form. Background Art

[0002] Traditional form design usually follows a static and predefined pattern, which means that at the beginning of the form creation, the designer needs to determine all the fields and layout. Although this method can provide a structured data processing method in some cases, it also has some obvious limitations, because the fixed nature of traditional forms limits its adaptability, and the preset form fields often cannot flexibly adapt to the changing user needs. Users have to fill in information in a series of fields that may not be related, which not only increases the burden on users, but may also lead to redundant and inaccurate information.

[0003] In some embodiments, modern form design begins to adopt more dynamic and intelligent methods. For example, by using conditional logic, the form can dynamically display or hide specific fields based on the user's input, thereby providing a highly targeted filling experience. However, this method requires field customization according to various filling situations, which requires a large amount of development work, and this form field display logic is also a pre-designed fixed field. Although it reduces the number of fields that users may need to fill in, it still cannot reflect the user's real filling needs. Summary of the invention

[0004] The present disclosure provides a method for generating dynamic fields in a form, wherein the form includes a plurality of fields, and the plurality of fields are selectively filled in by a user to complete the form filling; the method comprises:

[0005] Get the form type specified by the user;

[0006] Inputting the form type specified by the user into a trained machine learning model, so that the machine learning model performs association processing on the form type and field specified by the user to obtain the field corresponding to the form type specified by the user; wherein the machine learning model is obtained by supervised training based on field samples, and the field samples include the form type to which the field belongs;

[0007] Determining whether to trigger a dynamic field generation mechanism according to content filled in by the user in a field corresponding to the form type specified by the user;

[0008] If the dynamic field generation mechanism is triggered, the fields and filled-in contents filled in by the user are input into the trained machine learning model, so that the machine learning model performs association processing on the filled-in fields and filled-in contents, as well as the dynamic fields, to obtain predicted dynamic fields; wherein the machine learning model is obtained by supervised training based on the filled-in fields as training samples and the dynamic fields as labels, and the filled-in fields include the form type to which the fields belong, the fields, and the filled-in contents of the fields;

[0009] The predicted dynamic fields are returned to the user, so that the user can fill in the dynamic fields.

[0010] Optionally, the method further includes:

[0011] If the user fills in the preset content in the field corresponding to the form type specified by the user, or the user chooses to fill in the preset field in the field corresponding to the form type specified by the user, the dynamic field generation mechanism is triggered.

[0012] Optionally, after returning the predicted dynamic field to the user so that the user fills in the dynamic field, the method further includes:

[0013] The form type selected by the user is used as a newly added training sample. The user selects the filled-in fields in the fields corresponding to the form type specified by the user and the predicted dynamic fields as the labels of the newly added training samples, and the machine learning model is re-trained in a supervised manner.

[0014] Optionally, the form type selected by the user is used as a newly added training sample, the user selects a filled-in field from the fields corresponding to the form type specified by the user and the predicted dynamic fields as a label of the newly added training sample, and the machine learning model is re-trained in a supervised manner, including:

[0015] Receiving a custom field newly added by the user in addition to the field corresponding to the form type specified by the user and the predicted dynamic field;

[0016] The form type selected by the user is used as a newly added training sample. The user selects the filled-in fields from the fields corresponding to the form type specified by the user, the predicted dynamic fields and the newly added custom fields as the labels of the newly added training samples, and the machine learning model is re-trained in a supervised manner.

[0017] Optionally, after returning the predicted dynamic field to the user so that the user fills in the dynamic field, the method further includes:

[0018] The form type selected by the user, the fields and field contents selected and filled in by the user in the fields corresponding to the form type specified by the user are taken as new training samples, and the dynamic fields selected and filled in by the user in the predicted dynamic fields are taken as labels of the new training samples, and the machine learning model is re-trained in a supervised manner.

[0019] Optionally, the form type selected by the user, the fields and field contents selected by the user in the fields corresponding to the form type specified by the user and the predicted dynamic fields are used as newly added training samples, and the dynamic fields selected by the user in the predicted dynamic fields are used as labels of the newly added training samples, and the machine learning model is re-supervisedly trained, including:

[0020] Receiving a custom field newly added by the user in addition to the field corresponding to the form type specified by the user and the predicted dynamic field;

[0021] The form type selected by the user, the fields and field contents selected and filled in by the user in the fields corresponding to the form type specified by the user are taken as new training samples, and the dynamic fields selected and filled in by the user in the predicted dynamic fields and the newly added custom fields are taken as labels of the newly added training samples, and the machine learning model is re-trained in a supervised manner.

[0022] Optionally, returning the predicted dynamic field to the user so that the user fills in the dynamic field includes:

[0023] Returning the predicted dynamic field and a prompt identifier about the dynamic field to the user;

[0024] If the prompt mark of the dynamic field is clicked, prompt information is returned to the user to help the user understand that the dynamic field that appears is an applicable field predicted based on the same type of filling conditions in history, and fill in the dynamic field.

[0025] The present disclosure also provides a device for generating dynamic fields of a form, wherein the form includes a plurality of fields, and the plurality of fields are selectively filled by a user to complete the form filling; the device comprises:

[0026] The acquisition unit is used to obtain the form type specified by the user;

[0027] A first input unit is used to input the form type specified by the user into a trained machine learning model, so that the machine learning model performs association processing on the form type and field specified by the user to obtain the field corresponding to the form type specified by the user; wherein the machine learning model is obtained by supervised training based on field samples, and the field samples include the form type to which the field belongs;

[0028] A determination unit, configured to determine whether to trigger a dynamic field generation mechanism according to content filled in by a user in a field corresponding to the form type specified by the user;

[0029] The second input unit is used to input the fields and filled-in contents filled in by the user into the trained machine learning model if the dynamic field generation mechanism is triggered, so that the machine learning model performs association processing on the filled-in fields and filled-in contents and the dynamic fields to obtain predicted dynamic fields; wherein the machine learning model is obtained by supervised training based on the filled-in fields as training samples and the dynamic fields as labels, and the filled-in fields include the form type to which the fields belong, the fields, and the content filled in the fields;

[0030] The returning unit is used to return the predicted dynamic fields to the user so that the user can fill in the dynamic fields.

[0031] The present disclosure also provides an electronic device, comprising a communication interface, a processor, a memory and a bus, wherein the communication interface, the processor and the memory are interconnected via the bus;

[0032] The memory stores machine-readable instructions, and the processor executes the above method by calling the machine-readable instructions.

[0033] The present disclosure also provides a machine-readable storage medium, wherein the machine-readable storage medium stores machine-readable instructions, and when the machine-readable instructions are called and executed by a processor, the above method is implemented.

[0034] Through the embodiments of the present disclosure, after obtaining the form type specified by the user, the form type is input into the trained machine learning model, so that the machine learning model associates the form type and field specified by the user to obtain the field corresponding to the form type specified by the user; then, according to the content filled in by the user in the field corresponding to the form type specified by the user, it is determined whether to trigger the dynamic field generation mechanism. If triggered, the fields and content filled in by the user are input into the trained machine learning model, so that the machine learning model predicts the dynamic field; finally, the predicted dynamic field is returned to the user for the user to fill in. Accordingly, the present disclosure can realize the generation of targeted dynamic fields according to the fields and specific content selected by the user in the form, so that the user no longer needs to fill in redundant and irrelevant form field content, thereby improving the form filling efficiency and user adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions of the embodiments of this specification, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0036] Figure 1 is a schematic diagram of the architecture of a system for generating dynamic fields of a form according to an exemplary embodiment;

[0037] Figure 2 is a flowchart of a method for training a machine learning model shown in an exemplary embodiment;

[0038] Figure 3 It is a flow chart of a method for generating dynamic fields of a form shown in an exemplary embodiment;

[0039] Figure 4 is a flow chart of another method for generating dynamic fields of a form shown in an exemplary embodiment;

[0040] Figure 5 is a hardware structure diagram of an electronic device shown in an exemplary embodiment;

[0041] Figure 6 It is a block diagram of a device for generating dynamic fields of a form, showing an exemplary embodiment. DETAILED DESCRIPTION

[0042] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this specification.

[0043] It should be noted that: in other embodiments, the steps of the corresponding method are not necessarily performed in the order shown and described in this specification. In some other embodiments, the steps included in the method may be more or less than those described in this specification. In addition, a single step described in this specification may be decomposed into multiple steps for description in other embodiments; and multiple steps described in this specification may be combined into a single step for description in other embodiments.

[0044] Traditional form design usually follows a static and predefined pattern, which means that at the beginning of the form creation, the designer needs to determine all the fields and layout. Although this method can provide a structured data processing method in some cases, it also has some obvious limitations, because the fixed nature of traditional forms limits its adaptability, and the preset form fields often cannot flexibly adapt to the changing user needs. Users have to fill in information in a series of fields that may not be related, which not only increases the burden on users, but may also lead to redundant and inaccurate information.

[0045] For example, a traditional form for applying for a loan online may require users to provide a lot of personal information, such as name, address, proof of income, occupation details, credit score, etc. However, not every applicant needs to provide all of this information. For a long-term customer with a good credit record, the bank may not require them to provide detailed proof of income or work information; while for a new applicant, the bank may need more information for evaluation.

[0046] In this case, the traditional fixed form is somewhat rigid because it requires each applicant to fill in all fields, even if some information may be unnecessary. This not only increases the burden on users to fill in, but may also lead to inefficient data processing because back-end personnel need to filter out which information is relevant.

[0047] In some embodiments, modern form design begins to adopt more dynamic and intelligent methods. For example, by using conditional logic, the form can dynamically display or hide specific fields based on the user's input, thereby providing a highly targeted filling experience. However, this method requires field customization according to various filling situations, which requires a large amount of development work, and this form field display logic is also a pre-designed fixed field. Although it reduces the number of fields that users may need to fill in, it still cannot reflect the user's real filling needs.

[0048] In view of this, the present specification aims to propose a technical solution for generating highly targeted dynamic fields according to the situation of user filling in form fields.

[0049] During implementation, first, after obtaining the form type specified by the user, the form type is input into the trained machine learning model, so that the machine learning model associates the form type and fields specified by the user to obtain the fields corresponding to the form type specified by the user; further, based on the content filled in by the user in the fields corresponding to the form type specified by the user, it is determined whether to trigger the dynamic field generation mechanism. If triggered, the fields and content filled in by the user are input into the trained machine learning model, so that the machine learning model predicts the dynamic fields; finally, the predicted dynamic fields are returned to the user for the user to fill in.

[0050] For example, when a user opens a form filling application, the initial interface of the application has some default initial options, such as allowing the user to select the form type that he needs to fill in, which can be in the form of a dialogue or a drop-down option, etc., providing the user with several form types that can be selected, such as a product order form, a new user registration form, a job application form, a quarterly work report form, etc. Obtain the quarterly work report form specified by the user from the form list, and input the form type of the quarterly work report form into the trained machine learning model so that the machine learning model associates the form type of the quarterly work report form with its possibly related fields, and obtains the fields corresponding to the form type of the quarterly work report form, such as department name, project name, project progress, budget usage, problems encountered and suggestions, etc. During the filling process, if the user enters "80% completed" in the "project progress" field, or mentions "insufficient resources" in the "problems encountered and suggestions" field, the dynamic field generation mechanism will be triggered, and the machine learning model will predict the dynamic fields that may be needed, such as "detailed description of required resources" or "expected resource replenishment time" based on the user's input content and previous training data. Once the machine learning model predicts these dynamic fields, the application automatically adds them to the form and prompts the user to fill in further information. In this way, the user can provide more detailed information based on the predicted dynamic fields, such as filling in "Required resource detailed description" field "Need two additional developers and one server" and "Expected resource replenishment time" field "Expected within two weeks".

[0051] Finally, when the user completes filling in all fields, the application collects all information, including the original fields filled in by the user and the dynamic fields predicted by the model, so as to facilitate the subsequent re-supervised training of the machine learning model. This method not only improves the flexibility and personalization of form filling, but also helps to collect more comprehensive data, thereby providing stronger support for decision-making.

[0052] It can be seen that in the technical solution of this specification, after obtaining the form type specified by the user, the form type is input into the trained machine learning model so that the machine learning model associates the form type and field specified by the user to obtain the field corresponding to the form type specified by the user; then, according to the content filled in by the user in the field corresponding to the form type specified by the user, it is determined whether to trigger the dynamic field generation mechanism. If triggered, the fields and content filled in by the user are input into the trained machine learning model so that the machine learning model predicts the dynamic field; finally, the predicted dynamic field is returned to the user so that the user can fill it in. Accordingly, the present disclosure can realize the generation of targeted dynamic fields according to the fields and specific content selected by the user in the form, so that the user no longer needs to fill in redundant and irrelevant form field content, thereby improving the form filling efficiency and user adaptability.

[0053] The present disclosure is described below through specific embodiments in combination with specific application scenarios.

[0054] Please refer to Figure 1 , Figure 1 It is a schematic diagram of the architecture of a form dynamic field generation system shown in an exemplary embodiment of this specification.

[0055] like Figure 1 As shown, the above-mentioned form dynamic field generation system may include a server 11, a network 12, and several electronic devices (for example: mobile phone 13, mobile phone 14 and mobile phone 15).

[0056] The server 11 may be a server including an independent physical host, or a server cluster consisting of multiple independent physical hosts. Alternatively, the server 11 may be a virtual server, a cloud server, etc. carried by a host cluster. Mobile phones 13 to 15 are just one type of electronic device that a user can use. In fact, the user may also use electronic devices such as tablet devices, laptop computers, PDAs, etc., which are not limited in one or more embodiments of this specification. The network 12 may include various types of wired or wireless networks.

[0057] In actual applications, the application can run on the mobile phones 13 to 15, so that services can be provided externally based on the software and hardware on the mobile phones 13 to 15. It should be noted that the application can be an independent application or a small program that depends on an independent application, and this application does not limit this.

[0058] The above application can report the form type selected by the user, the form fields selected by the user and their contents to the server. The server can be deployed on the server 11. The server can associate fields with the form type selected by the user, thereby providing the user with fields corresponding to the form type selected by the user; the server can also parse the form fields selected by the user and their contents to determine whether to trigger the dynamic field generation mechanism, and if triggered, associate the form fields selected by the user and their contents with the predicted dynamic fields.

[0059] The technical solution provided by one or more embodiments of this specification involves using a pre-trained machine learning model to associate fields with the form type selected by the user, thereby providing the user with fields corresponding to the form type selected by the user; the form fields and their contents selected by the user to fill in can also be parsed to determine whether to trigger the dynamic field generation mechanism, and if triggered, the form fields and their contents selected by the user to fill in are associated with the predicted dynamic fields. In order to facilitate the understanding of the technical solution in this specification, the training process of the machine learning model will be described first, and then the use process of the machine learning model will be described.

[0060] 1. Training process of machine learning model

[0061] Please combine Figure 1 ,refer to Figure 2 , Figure 2 It is a flowchart of a training method for a machine learning model shown in an exemplary embodiment of this specification.

[0062] In this embodiment, the above-mentioned form dynamic field generation method can be applied to the following Figure 1 The server side of the form dynamic field generation system shown.

[0063] The training method of the above machine learning model may include the following steps:

[0064] Step 202: Obtain field samples of filled-in fields in a form actually filled out by the user; wherein the field samples are marked with the form type to which the field belongs and the associated dynamic field.

[0065] In order to train the machine learning model, field samples of filled fields in several forms actually filled out by users can be obtained first. For example, the server can store the obtained filled forms (for example, the filled forms reported by the applications of mobile phones 13 to 15) as the forms actually filled out by users, so as to directly obtain field samples of filled fields in several forms actually filled out by users.

[0066] For any form actually filled out by a user, the field samples of the filled-in fields in the form actually filled out by the user are marked with the form type to which the fields belong.

[0067] Continuing with the above machine learning model as an example, the field samples of the filled-in fields in the form that the user actually filled out (for example, department name, project name, project progress) are marked with the form type to which the field belongs (for example, quarterly work report form) and the associated dynamic fields (for example, detailed description of required resources).

[0068] Step 204: Using the field samples that are labeled with the form type to which the field belongs and the associated dynamic fields as training samples, supervised training is performed on the machine learning model to obtain the machine learning model.

[0069] In this embodiment, when the field samples of the form types and associated dynamic fields to which the fields are labeled are obtained, since the field samples have been labeled, the machine learning model can be supervisedly trained based on the field samples of the form types and associated dynamic fields to which the fields are labeled. For example, the field samples of the form types and associated dynamic fields to which the fields are labeled can be input into the machine learning model, and the machine learning model performs association learning on the field samples of the form types and associated dynamic fields to which the fields are labeled, and obtains the prediction results for the fields corresponding to the form types, and the prediction results for the dynamic fields predicted for the filled-in field contents. Subsequently, based on the two prediction results, the losses can be calculated for the annotations of the form types contained in each sample field, and for the annotations of the associated dynamic fields contained in each sample field, and the model parameters of the machine learning model can be adjusted according to the calculated losses; until the loss reaches the minimum value, the machine learning model training can be considered completed. At this point, the machine learning model is trained.

[0070] In the technical solution of the present disclosure, since the Naive Bayesian model has shown good performance in the field of text classification, the algorithm principle is simple, it is suitable for incremental training, and it is less dependent on training data, the Naive Bayesian model can be used as a machine learning model to apply the present solution. The present disclosure does not limit the specific types of machine learning models.

[0071] 2. The use process of machine learning models

[0072] Please combine Figure 1 ,refer to Figure 3 , Figure 3 The figure is a flowchart of a method for generating dynamic fields of a form, shown in an exemplary embodiment.

[0073] In this embodiment, the above-mentioned form dynamic field generation method can be applied to the following Figure 1 The server side of the form dynamic field generation system shown.

[0074] The above-mentioned form dynamic field generation method may include the following steps:

[0075] Step 302: Obtain the form type specified by the user.

[0076] For example, when a user opens an application for filling out forms, the initial interface of the application has some default initial options, such as allowing the user to select the form type they need to fill out. The specific form can be a dialogue or a drop-down option, etc., providing the user with several selectable form types, such as a product order form, a new user registration form, a job application form, a quarterly work report form, etc. Get the quarterly work report form specified by the user from the form list.

[0077] Among them, form types refer to different categories or uses of forms. Each form type contains a specific set of fields designed to collect specific types of data. For example, a product order form will contain fields such as product name, product quantity, and delivery address, while a new user registration form will contain fields such as username, password, and email address. The initial interface refers to the first screen or page that a user sees when they first launch an application or visit a website. It is usually where users are guided to take the next step. The conversational interface is a way of human-computer interaction that can communicate with users in natural language in the form of a chatbot or virtual assistant to understand users' needs and guide users to choose the appropriate form type. The drop-down option is a common UI element. When a user clicks it, a drop-down menu will appear, listing a series of items to choose from. The user can directly click or check the required form type.

[0078] Step 304: Input the form type specified by the user into the trained machine learning model, so that the machine learning model can associate the form type and field specified by the user to obtain the field corresponding to the form type specified by the user; wherein the machine learning model is obtained by supervised training based on field samples, and the field samples include the form type to which the field belongs.

[0079] For example, the form type of quarterly work report form specified by the user from the form list is input into the trained machine learning model, so that the machine learning model can associate the form type of quarterly work report form with its possibly related fields, and obtain the fields corresponding to the form type of quarterly work report form, such as department name, project name, project progress, budget usage, problems encountered and suggestions, etc.

[0080] Among them, supervised training is a training method for machine learning, where the machine learning model learns from labeled training data. In supervised training, each training sample has a corresponding output label, and the training goal of the machine learning model is to learn how to predict these labels based on the input data. Before training the machine learning model, data preprocessing can be performed, such as cleaning, formatting, and encoding field samples so that the machine learning model can understand them correctly.

[0081] Step 306: Determine whether to trigger a dynamic field generation mechanism according to the content filled in by the user in the field corresponding to the form type specified by the user.

[0082] For example, when the user is filling in the form, if he / she enters "80% completed" in the "Project Progress" field, or mentions "Insufficient resources" in the "Problems encountered and suggestions" field, the dynamic field generation mechanism will be triggered.

[0083] The trigger condition is a preset rule condition. When the user input meets the preset rule condition, the dynamic field generation mechanism will be activated. The dynamic field generation mechanism is a method of dynamically adding or adjusting form fields based on the user input. It allows new input fields to automatically appear based on specific keywords or phrases during the user filling process in order to collect more relevant information. When implementing this dynamic field generation mechanism, developers need to consider the user experience and ensure that the appearance of new fields is timely, relevant, and does not cause trouble or confusion to users. In addition, it is also necessary to ensure that the system's response time is fast enough to provide users with a smooth filling experience.

[0084] Step 308: If the dynamic field generation mechanism is triggered, the fields and contents filled in by the user are input into the trained machine learning model, so that the machine learning model can associate the filled in fields and contents, as well as the dynamic fields, to obtain predicted dynamic fields; wherein the machine learning model is obtained through supervised training based on the filled in fields as training samples and the dynamic fields as labels, and the filled in fields include the form type to which the field belongs, the field, and the content filled in the field.

[0085] For example, based on the user's input of "80% completed" in the "Project Progress" field or "Insufficient resources" in the "Problems and Suggestions" field while filling out the form, the machine learning model will combine previous training data to predict dynamic fields that may be needed, such as "Detailed description of required resources" or "Expected resource replenishment time."

[0086] Among them, the above-mentioned dynamic fields are dynamic fields related to the filled-in fields inferred by the trained machine learning model based on the fields and contents filled in the form by the user. These dynamic fields can help the form maker to further obtain more detailed information, and can also help the users who fill in the form to feel a better targeted and personalized experience in the process of filling out the form. It can help to provide complete and accurate relevant information at one time and avoid wasting energy on irrelevant fields.

[0087] Step 310: Return the predicted dynamic fields to the user, so that the user can fill in the dynamic fields.

[0088] For example, once the machine learning model predicts these dynamic fields, the application automatically adds them to the form and prompts the user to fill in further information. In this way, the user can provide more detailed information based on the predicted dynamic fields, such as filling in "Required resource detailed description" field "Need two additional developers and one server" and "Expected resource replenishment time" field "Expected within two weeks".

[0089] Among them, after the dynamic fields are generated, in order to facilitate users to enter relevant information in these newly appeared dynamic fields, it usually involves updating the user interface (UI) and may also involve dynamically creating form controls, such as text boxes, drop-down menus, etc., to ensure that the new fields can be seen and interacted with by users.

[0090] In one embodiment shown, the method further includes: if the user fills in preset content in the field corresponding to the form type specified by the user, or the user chooses to fill in a preset field in the field corresponding to the form type specified by the user, then a dynamic field generation mechanism is triggered.

[0091] For example, when filling in the form, if the user enters "80% completed", "Not completed", etc. in the "Project progress" field, or mentions "Insufficient resources" in the "Problems encountered and suggestions" field, the dynamic field generation mechanism will be triggered.

[0092] The present specification does not limit the specific conditions and settings for triggering the dynamic field generation mechanism.

[0093] In one embodiment shown, after returning the predicted dynamic fields to the user so that the user fills in the dynamic fields, the method further includes: taking the form type selected by the user as a newly added training sample, the user selects the filled-in fields from the fields corresponding to the form type specified by the user and the predicted dynamic fields as the labels of the newly added training samples, and re-performing supervised training on the machine learning model.

[0094] For example, after the user has filled in the predicted dynamic fields, including "detailed description of required resources", "expected resource replenishment time" and "other notes on expected resources", the form type selected by the user (quarterly work report form) is used as a training sample, and the user selects the filled-in fields ("department name", "project name", "project progress", "budget usage" and "problems encountered and suggestions") in the fields corresponding to the form type of the quarterly work report form and the predicted dynamic fields ("detailed description of required resources", "expected resource replenishment time" and "other notes on expected resources") as the labels of the newly added training samples, and the machine learning model is retrained. The trained machine learning model will obtain the fields corresponding to the form type the next time the user selects the form type as the quarterly work report form, and the probability of the occurrence of the "detailed description of required resources" and "expected resource replenishment time" fields is increased, and the probability of the occurrence of the "budget usage" and "problems encountered and suggestions" fields is reduced.

[0095] Among them, when a user fills out the "quarterly work report form" and selects specific fields to fill in, this information is used as a training sample to retrain the machine learning model. The machine learning model will learn the user's behavior patterns when filling out a specific form type and adjust its predictions based on these patterns. In the future, when other users select the same form type, the model will be able to more accurately predict which fields are more likely to be filled in, thereby improving the fields corresponding to the form type to improve user experience and data collection efficiency.

[0096] In one embodiment shown, the form type selected by the user is used as a newly added training sample, the user selects the filled-in fields from the fields corresponding to the form type specified by the user and the predicted dynamic fields as the labels of the newly added training samples, and the machine learning model is re-supervised trained, including: receiving a newly added custom field by the user in addition to the fields corresponding to the form type specified by the user and the predicted dynamic fields; the form type selected by the user is used as a newly added training sample, the user selects the filled-in fields from the fields corresponding to the form type specified by the user, the predicted dynamic fields and the newly added custom fields as the labels of the newly added training samples, and the machine learning model is re-supervised trained.

[0097] For example, a user-added custom field "expected resource support department" is received, which is neither a field corresponding to the form type specified by the user nor a predicted dynamic field. In this case, the form type selected by the user (quarterly work report form) is used as a training sample, and the user selects the filled fields ("department name", "project name", "project progress", "budget usage" and "problems encountered and suggestions") in the fields corresponding to the quarterly work report form form type, the predicted dynamic fields ("required resource detailed description", "expected resource replenishment time" and "expected resource other notes") and the newly added custom field ("expected resource support department") as the labels of the newly added training samples, and the machine learning model is retrained. The trained machine learning model obtains that the probability of the fields "required resource detailed description", "expected resource replenishment time" and "expected resource support department" in the fields corresponding to the form type is increased, and the probability of the fields "budget usage" and "problems encountered and suggestions" is reduced.

[0098] Custom fields are fields added to the form by users based on specific needs. They are not predefined form fields or dynamic fields generated by system predictions. Users can add new fields to collect specific information based on their needs. During the retraining of the machine learning model, the model will learn that users are likely to fill in this custom field in this type of form, so the probability of this field appearing in future predictions will increase.

[0099] In this embodiment, this training method helps the model better adapt to the specific needs of users and provide a more personalized form filling experience. This allows the form system to not only adapt to predefined needs, but also flexibly adapt to user-specific, unforeseen needs, thereby improving the flexibility and effectiveness of data collection.

[0100] In one embodiment shown, after returning the predicted dynamic fields to the user so that the user fills in the dynamic fields, the method further includes: using the form type selected by the user, the fields and field contents selected by the user to fill in the fields corresponding to the form type specified by the user as newly added training samples, and the dynamic fields selected by the user to fill in the predicted dynamic fields as labels of the newly added training samples, and re-performing supervised training on the machine learning model.

[0101] For example, in the fields corresponding to the form type (quarterly work report form) selected by the user, namely, "department name", "project name", "project progress", "budget usage" and "problems encountered and suggestions", except for the "budget usage" and "problems encountered and suggestions" fields, the other fields are filled in. In the predicted dynamic fields, including "detailed description of required resources", "expected resource replenishment time" and "other notes for expected resources", except for the "other notes for expected resources" field, the other dynamic fields are filled in. Then, the form type (quarterly work report form) selected by the user, the fields selected by the user ("detailed description of required resources", "expected resource replenishment time", "department name", "project name", "project progress") and the field contents can be used as new training samples to retrain the machine learning model. After the training, the next time the user selects the form type as the quarterly work report form, the appearance frequency of the "budget usage" and "problems encountered and suggestions" fields in the corresponding fields will decrease. When the dynamic field generation mechanism is triggered under the same filling condition, the appearance frequency of the "other notes for expected resources" field will decrease.

[0102] In machine learning, training samples are data points used to train machine learning models. In this embodiment, the form fields selected and filled in by the user and the specific field contents will become new training data for further training and improving the machine learning model. This may involve adjusting model parameters and re-optimizing the model to improve the prediction accuracy of the machine learning model, so that the machine learning model can continue to learn from new user interactions over time to adapt to changes in user behavior.

[0103] It should be noted that in the scheme disclosed herein, fields corresponding to the form type and which fields in the predicted dynamic fields are required can be set as needed. These fields are information that must be collected, and there is no need to consider whether asking users to fill them in will cause redundancy and trouble to users. These fields are also not within the scope of machine learning model learning and training.

[0104] In one embodiment shown, the form type selected by the user, the fields selected and filled in by the user in the fields corresponding to the form type specified by the user and the predicted dynamic fields are used as newly added training samples, the dynamic fields selected and filled in by the user in the predicted dynamic fields are used as labels of the newly added training samples, and the machine learning model is re-supervised trained, including: receiving a custom field newly added by the user in addition to the fields corresponding to the form type specified by the user and the predicted dynamic fields; the form type selected by the user, the fields selected and filled in by the user in the fields corresponding to the form type specified by the user and the field contents are used as newly added training samples, the dynamic fields selected and filled in by the user in the predicted dynamic fields and the newly added custom fields are used as labels of the newly added training samples, and the machine learning model is re-supervised trained.

[0105] For example, a user-added custom field "Expected Resource Support Department" is received. This field does not belong to the fields corresponding to the form type specified by the user, nor to the predicted dynamic fields. In this case, the form type selected by the user (quarterly work report form) and the fields corresponding to the form type, namely "department name", "project name", "project progress", "budget usage" and "problems and suggestions encountered", the fields "department name", "project name" and "project progress" selected and filled in, and the field contents are used as the newly added training samples; the fields "required resource detailed description", "expected resource replenishment time" and "expected resource support department" selected and filled in by the user in the predicted dynamic fields and the newly added custom fields, and the field contents are used as the labels of the newly added training samples, and the machine learning model is re-trained in a supervised manner. After the training, the frequency of the "budget usage" and "problems and suggestions encountered" fields in the corresponding fields will decrease when the user selects the form type as the quarterly work report form next time. When the dynamic field generation mechanism is triggered under the same filling condition, the frequency of the "expected resource other notes" field will decrease, and the frequency of the "expected resource support department" field will increase.

[0106] It should be noted that the field corresponding to the form type specified by the user is the most relevant field, and the predicted dynamic field belongs to the field that the user predicts may be relevant when the content filled in belongs to a specific situation. The frequency of use of these two fields varies. Although the user-defined field can theoretically increase the probability of the user-defined field appearing in the field corresponding to the form type specified by the user, and increase the probability of the user-defined field appearing in the predicted dynamic field under the same filling condition, considering that the user-defined field is a more personalized choice, in the process of using the custom field to train the machine learning model, as the frequency of filling in the custom field increases, it will first be reflected in the predicted dynamic field that the custom field is more likely to appear. In the predicted dynamic field, the increase in the frequency of the custom field being selected to be filled in may be reflected in the field corresponding to the form type specified by the user. Regarding the weight selection of the machine learning model for these two fields, and the specific training method, this manual does not limit this.

[0107] In one embodiment shown, the returning the predicted dynamic field to the user so that the user can fill in the dynamic field includes: returning the predicted dynamic field and a prompt mark about the dynamic field to the user; if the prompt mark of the dynamic field is clicked, returning prompt information to the user to help the user understand that the dynamic field that appears is an applicable field predicted based on the same type of filling conditions in history, and fill in the dynamic field.

[0108] For example, a user is filling out a "quarterly work report form", and during the filling process, the dynamic field generation mechanism is triggered. The machine learning model predicts that the user may need to provide additional information about resource requirements, so two dynamic fields are generated: "Detailed description of required resources" and "Expected resource replenishment time". There will be a prompt icon next to these dynamic fields (for example, a small question mark icon). When the user hovers over or clicks this icon, a prompt box or sidebar will pop up with the following information: "Based on the filling of similar quarterly work reports in the past, we found that detailed descriptions of resource requirements and resource replenishment time are common requirements. Please provide the type and quantity of resources you need in the 'Detailed description of required resources' field, and when you expect these resources to be in place in the 'Expected resource replenishment time' field."

[0109] In this embodiment, such prompt information helps users understand why these fields appear in the form and how to fill them in. This interactive mode not only improves the user experience, but also ensures that the collected data is more complete and accurate.

[0110] This specification does not limit the form and specific content of the prompt information.

[0111] In order to help those skilled in the art better understand this method, Figure 4 This method is described.

[0112] See also Figure 4 , Figure 4 FIG. 1 is a flowchart of another method for generating dynamic fields of a form according to an exemplary embodiment. Figure 4 As shown, when the method is started, the form type specified by the user is first obtained, the fields corresponding to the form type are obtained, and a form is generated based on these fields for the user to fill in; then the fields and contents selected by the user in the form fields are received, and based on this information, it is determined whether to trigger the dynamic field generation mechanism. If not triggered, the form filled in and submitted by the user is directly stored, and the form data filled in this time is used as a new sample to re-train the machine learning model in a supervised manner; if triggered, the machine learning model is called to predict the dynamic fields based on the form fields and their contents currently filled in by the user, assemble the dynamic form data, and generate the form part with dynamic fields added for the user to continue filling in.

[0113] Corresponding to the above-mentioned embodiment of the method for generating dynamic fields in a form, this specification also provides an embodiment of a device for generating dynamic fields in a form.

[0114] See also Figure 5 , Figure 5 It is a hardware structure diagram of an electronic device shown in an exemplary embodiment. At the hardware level, the device includes a processor 502, an internal bus 504, a network interface 506, a memory 508 and a non-volatile memory 510, and of course may also include other required hardware. One or more embodiments of this specification can be implemented based on software, such as the processor 502 reading the corresponding computer program from the non-volatile memory 510 into the memory 508 and then running it. Of course, in addition to the software implementation, one or more embodiments of this specification do not exclude other implementation methods, such as logic devices or a combination of software and hardware, etc., that is to say, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.

[0115] See also Figure 6 , Figure 6 FIG. 1 is a block diagram of a form dynamic field generation device according to an exemplary embodiment. The form dynamic field generation device can be applied to Figure 5 In the electronic device shown, the technical solution of this specification is implemented. The form includes several fields, and the several fields are selectively filled by the user to complete the form filling; the device includes:

[0116] An acquisition unit 602 is used to acquire a form type specified by a user;

[0117] A first input unit 604 is used to input the form type specified by the user into a trained machine learning model, so that the machine learning model performs association processing on the form type and field specified by the user to obtain the field corresponding to the form type specified by the user; wherein the machine learning model is obtained by supervised training based on field samples, and the field samples include the form type to which the field belongs;

[0118] A determination unit 606, configured to determine whether to trigger a dynamic field generation mechanism according to the content filled in by the user in the field corresponding to the form type specified by the user;

[0119] The second input unit 608 is used to input the fields and filled-in contents of the user into the trained machine learning model if the dynamic field generation mechanism is triggered, so that the machine learning model performs association processing on the filled-in fields and filled-in contents and the dynamic fields to obtain predicted dynamic fields; wherein the machine learning model is obtained by supervised training based on the filled-in fields as training samples and the dynamic fields as labels, and the filled-in fields include the form type to which the fields belong, the fields, and the filled-in contents of the fields;

[0120] The returning unit 610 is configured to return the predicted dynamic fields to the user, so that the user can fill in the dynamic fields.

[0121] In this embodiment, the device further includes:

[0122] The trigger unit is used to trigger the dynamic field generation mechanism if the user fills in the preset content in the field corresponding to the form type specified by the user, or the user chooses to fill in the preset field in the field corresponding to the form type specified by the user.

[0123] In this embodiment, after returning the predicted dynamic field to the user so that the user fills in the dynamic field, the apparatus further includes:

[0124] The first training unit is used to use the form type selected by the user as a newly added training sample. The user selects the filled-in field from the fields corresponding to the form type specified by the user and the predicted dynamic fields as the label of the newly added training sample, and re-performs supervised training on the machine learning model.

[0125] In this embodiment, the first training unit includes:

[0126] A first receiving subunit is used to receive a custom field added by the user in addition to the field corresponding to the form type specified by the user and the predicted dynamic field;

[0127] The first training sub-unit is used to use the form type selected by the user as a newly added training sample. The user selects the filled-in fields from the fields corresponding to the form type specified by the user, the predicted dynamic fields and the newly added custom fields as the labels of the newly added training samples, and re-performs supervised training on the machine learning model.

[0128] In this embodiment, after returning the predicted dynamic field to the user so that the user fills in the dynamic field, the apparatus further includes:

[0129] The second training unit is used to use the form type selected by the user, the fields and field contents selected and filled in by the user in the fields corresponding to the form type specified by the user as new training samples, and the dynamic fields selected and filled in by the user in the predicted dynamic fields as labels of the new training samples, and re-perform supervised training on the machine learning model.

[0130] In this embodiment, the second training unit includes:

[0131] A second receiving subunit is used to receive a custom field added by the user in addition to the field corresponding to the form type specified by the user and the predicted dynamic field;

[0132] The second training sub-unit is used to use the form type selected by the user, the fields and field contents selected and filled in by the user in the fields corresponding to the form type specified by the user as newly added training samples, and the dynamic fields selected and filled in by the user in the predicted dynamic fields and the newly added custom fields as labels of the newly added training samples, and re-perform supervised training on the machine learning model.

[0133] In this embodiment, the return unit includes:

[0134] A first returning subunit, configured to return the predicted dynamic field and a prompt mark about the dynamic field to the user;

[0135] The second return subunit is used to return prompt information to the user if the prompt mark of the dynamic field is clicked, so as to help the user understand that the dynamic field that appears is an applicable field predicted based on the historical filling conditions of the same type, and fill in the dynamic field.

[0136] The implementation process of the functions and effects of each unit in the above-mentioned device is specifically described in the implementation process of the corresponding steps in the above-mentioned method, and will not be repeated here.

[0137] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can refer to the partial description of the method embodiments. The device embodiments described above are only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this specification. Ordinary technicians in this field can understand and implement it without paying creative work.

[0138] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer, which may be in the form of a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email transceiver, a game console, a tablet computer, a wearable device or a combination of any of these devices.

[0139] In a typical configuration, a computer includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0140] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0141] Computer-readable media include permanent and non-permanent, removable and non-removable media that can be used to store information by any method or technology. Information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, disk storage, quantum memory, graphene-based storage media or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include temporary computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0142] The user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this disclosure are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0143] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device 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, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0144] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0145] The terms used in one or more embodiments of this specification are only for the purpose of describing specific embodiments, and are not intended to limit one or more embodiments of this specification. The singular forms of "one", "said" and "the" used in one or more embodiments of this specification and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0146] It should be understood that although the terms first, second, third, etc. may be used to describe various information in one or more embodiments of this specification, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of one or more embodiments of this specification, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0147] The above description is merely a preferred embodiment of one or more embodiments of the present specification and is not intended to limit one or more embodiments of the present specification. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of one or more embodiments of the present specification shall be included in the scope of protection of one or more embodiments of the present specification.

Claims

1. A method for generating dynamic fields in a form, characterized in that: The form includes a plurality of fields, and the plurality of fields are selectively filled by the user to complete the form filling; the method includes: Get the form type specified by the user; Inputting the form type specified by the user into a trained machine learning model, so that the machine learning model performs association processing on the form type and field specified by the user to obtain the field corresponding to the form type specified by the user; wherein the machine learning model is obtained by supervised training based on field samples, and the field samples include the form type to which the field belongs; Determining whether to trigger a dynamic field generation mechanism according to content filled in by the user in a field corresponding to the form type specified by the user; If the dynamic field generation mechanism is triggered, the fields and filled-in contents filled in by the user are input into the trained machine learning model, so that the machine learning model performs association processing on the filled-in fields and filled-in contents, as well as the dynamic fields, to obtain predicted dynamic fields; wherein the machine learning model is obtained by supervised training based on the filled-in fields as training samples and the dynamic fields as labels, and the filled-in fields include the form type to which the fields belong, the fields, and the filled-in contents of the fields; The predicted dynamic fields are returned to the user, so that the user can fill in the dynamic fields.

2. The method according to claim 1, characterized in that: The method further comprises: If the user fills in the preset content in the field corresponding to the form type specified by the user, or the user chooses to fill in the preset field in the field corresponding to the form type specified by the user, the dynamic field generation mechanism is triggered.

3. The method according to claim 1, characterized in that After returning the predicted dynamic fields to the user so that the user can fill in the dynamic fields, the method further comprises: The form type selected by the user is used as a newly added training sample. The user selects the filled-in fields in the fields corresponding to the form type specified by the user and the predicted dynamic fields as the labels of the newly added training samples, and the machine learning model is re-trained in a supervised manner.

4. The method according to claim 3, characterized in that The method of using the form type selected by the user as a newly added training sample, wherein the user selects a filled-in field from among the fields corresponding to the form type specified by the user and the predicted dynamic fields as a label of the newly added training sample, and re-performing supervised training on the machine learning model includes: Receiving a custom field newly added by the user in addition to the field corresponding to the form type specified by the user and the predicted dynamic field; The form type selected by the user is used as a newly added training sample. The user selects the filled-in fields from the fields corresponding to the form type specified by the user, the predicted dynamic fields and the newly added custom fields as the labels of the newly added training samples, and the machine learning model is re-trained in a supervised manner.

5. The method according to claim 1, characterized in that After returning the predicted dynamic fields to the user so that the user can fill in the dynamic fields, the method further comprises: The form type selected by the user, the fields and field contents selected and filled in by the user in the fields corresponding to the form type specified by the user are taken as new training samples, and the dynamic fields selected and filled in by the user in the predicted dynamic fields are taken as labels of the new training samples, and the machine learning model is re-trained in a supervised manner.

6. The method according to claim 5, characterized in that The method further comprises: using the form type selected by the user, the fields and field contents selected by the user in the fields corresponding to the form type specified by the user and the predicted dynamic fields as newly added training samples, and selecting the dynamic fields filled in by the user in the predicted dynamic fields as labels of the newly added training samples, and re-performing supervised training on the machine learning model, including: Receiving a custom field newly added by the user in addition to the field corresponding to the form type specified by the user and the predicted dynamic field; The form type selected by the user, the fields and field contents selected and filled in by the user in the fields corresponding to the form type specified by the user are taken as new training samples, and the dynamic fields selected and filled in by the user in the predicted dynamic fields and the newly added custom fields are taken as labels of the newly added training samples, and the machine learning model is re-trained in a supervised manner.

7. The method according to claim 1, characterized in that The returning the predicted dynamic field to the user so that the user fills in the dynamic field comprises: Returning the predicted dynamic field and a prompt identifier about the dynamic field to the user; If the prompt mark of the dynamic field is clicked, prompt information is returned to the user to help the user understand that the dynamic field that appears is an applicable field predicted based on the same type of filling conditions in history, and fill in the dynamic field.

8. A device for generating dynamic fields in a form, characterized in that: The form includes a plurality of fields, and the plurality of fields are selectively filled by the user to complete the form filling; the device includes: The acquisition unit is used to obtain the form type specified by the user; A first input unit is used to input the form type specified by the user into a trained machine learning model, so that the machine learning model performs association processing on the form type and field specified by the user to obtain the field corresponding to the form type specified by the user; wherein the machine learning model is obtained by supervised training based on field samples, and the field samples include the form type to which the field belongs; A determination unit, configured to determine whether to trigger a dynamic field generation mechanism according to content filled in by a user in a field corresponding to the form type specified by the user; The second input unit is used to input the fields and filled-in contents filled in by the user into the trained machine learning model if the dynamic field generation mechanism is triggered, so that the machine learning model performs association processing on the filled-in fields and filled-in contents and the dynamic fields to obtain predicted dynamic fields; wherein the machine learning model is obtained by supervised training based on the filled-in fields as training samples and the dynamic fields as labels, and the filled-in fields include the form type to which the fields belong, the fields, and the content filled in the fields; The returning unit is used to return the predicted dynamic fields to the user so that the user can fill in the dynamic fields.

9. An electronic device, characterized in that: It includes a communication interface, a processor, a memory and a bus, wherein the communication interface, the processor and the memory are interconnected via the bus; The memory stores machine-readable instructions, and the processor executes the method according to any one of claims 1 to 7 by calling the machine-readable instructions.

10. A machine-readable storage medium, characterized in that: The machine-readable storage medium stores machine-readable instructions, and when the machine-readable instructions are called and executed by the processor, the method according to any one of claims 1 to 7 is implemented.

Images