Form content prediction method, device, electronic device and storage medium

Abstract

The embodiments of the present invention are applicable to the field of data processing technology, and provide a form content prediction method, device, electronic device and storage medium, wherein the form content prediction method includes: receiving a filling request; the filling request is used to request filling in a form; in the case where the filling request is associated with at least two forms, determining whether the at least two forms include at least two forms with a chain relationship; the chain relationship represents that the at least two forms have a filling order, and the filling content of the first form is determined based on the filling content of the second form; the filling order corresponding to the first form is located before the filling order corresponding to the second form; in the case where the at least two forms include at least two forms with a chain relationship, predicting the filling content of the at least two forms with a chain relationship based on a set model cluster; each form of the at least two forms with a chain relationship corresponds to a model in the set model cluster.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a form content prediction method, device, electronic device and storage medium. Background Art

[0002] In the e-commerce field, when users purchase goods or return or exchange goods, they will generate form information filling requests caused by orders. For example, when users apply for after-sales service, they will generate express delivery information filling requests and after-sales information filling requests. Related technologies use deep learning models to predict the content of form filling. Deep learning models only support single-round interaction, that is, they can only predict the content of one form at a time, and lack multi-round interaction capabilities, that is, the ability to predict the content of multiple forms at a time. Summary of the invention

[0003] In order to solve the above problems, embodiments of the present invention provide a form content prediction method, device, electronic device and storage medium, so as to at least solve the problem that the related art cannot predict the filled-in contents of multiple forms at one time.

[0004] The technical solution of the present invention is achieved in this way:

[0005] In a first aspect, an embodiment of the present invention provides a form content prediction method, the method comprising:

[0006] Receive a filling request; the filling request is used to request filling in a form;

[0007] In the case where the filling request is associated with at least two forms, determining whether the at least two forms include at least two forms having a chain relationship; the chain relationship indicates that the at least two forms have a filling sequence, and the filling content of the first form is determined based on the filling content of the second form; and the filling sequence corresponding to the first form is before the filling sequence corresponding to the second form;

[0008] In the case where the at least two forms include at least two forms with a chain relationship, the filled-in contents of the at least two forms with a chain relationship are predicted based on a set model cluster; each form of the at least two forms with a chain relationship corresponds to a model in the set model cluster.

[0009] In the above solution, when predicting the filled-in contents of the at least two forms having a chain relationship based on the set model cluster, the method includes:

[0010] The filled-in content of the first form is input into the first model corresponding to the second form in the set model cluster to obtain the filled-in content of the second form predicted by the corresponding first model.

[0011] In the above scheme, the method further includes:

[0012] In the case where the filling request is associated with only one form, the filling content of the form is predicted based on the set second model; the set second model is used to predict the filling content of the form associated with the filling request associated with only one form.

[0013] In the above scheme, the method further includes:

[0014] In the case where the at least two forms do not include at least two forms with a chain relationship, the filled-in content of each of the at least two forms is predicted based on a set third model; the set third model is used to predict the filled-in content of the at least two forms that do not include the at least two forms with a chain relationship.

[0015] In the above scheme, the method further includes:

[0016] Monitoring at least one indicator corresponding to each model in at least one model;

[0017] When the value of any one of the at least one indicator is less than a set value, updating the corresponding model;

[0018] The at least one model includes at least one of the following:

[0019] All models in the set model cluster;

[0020] The second model of setting;

[0021] The third model of the setting.

[0022] In the above solution, the at least one indicator includes at least one of the following:

[0023] Accuracy rate; the accuracy rate represents the ratio of the number of correct predictions made by the corresponding model within a set time period to the total number of predictions;

[0024] Precision rate; the precision rate represents the ratio of the number of times the corresponding model correctly predicts the form's filled-in content to the number of times the model correctly predicts the form's category within a set time period;

[0025] Adoption rate; the adoption rate represents the proportion of the number of times the form content is filled in the form out of the number of times the corresponding model predicts correctly within a set time period;

[0026] Assistance rate; the assistance rate represents the ratio of the number of filled-in forms among the filled-in forms predicted correctly by the corresponding model to the total number of forms.

[0027] In the above scheme, the method further includes:

[0028] Determine a form corresponding to the filled-in content of the form in a setting database based on the set form items; the filled-in content of the form includes the set form items; the setting database is used to store forms;

[0029] Fill in the form contents into the corresponding form.

[0030] In the above solution, the set form item includes any one of the following:

[0031] Order ID;

[0032] Express delivery note identification;

[0033] User ID.

[0034] In a second aspect, an embodiment of the present invention provides a form content prediction device, the device comprising:

[0035] A receiving module, used to receive a filling request; the filling request is used to request to fill in a form;

[0036] A determination module, configured to determine, when the filling request is associated with at least two forms, whether the at least two forms include at least two forms having a chain relationship; the chain relationship indicates that the at least two forms have a filling sequence, and wherein the filling content of the first form is determined based on the filling content of the second form; and the filling sequence corresponding to the first form is before the filling sequence corresponding to the second form;

[0037] A prediction module is used to predict the filled-in contents of the at least two forms with a chain relationship based on a set model cluster when the at least two forms include at least two forms with a chain relationship; each form in the at least two forms corresponds to a model in the set model cluster.

[0038] In a third aspect, an embodiment of the present invention provides an electronic device, comprising a processor and a memory, wherein the processor and the memory are interconnected, wherein the memory is used to store a computer program, the computer program comprises program instructions, and the processor is configured to call the program instructions to execute the steps of the form content prediction method provided in the first aspect of the embodiment of the present invention.

[0039] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, including: the computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the steps of the form content prediction method provided in the first aspect of the embodiment of the present invention are implemented.

[0040] The embodiment of the present invention receives a filling request, and in the case where the filling request is associated with at least two forms, determines whether the at least two forms include at least two forms with a chain relationship, and in the case where the at least two forms include at least two forms with a chain relationship, predicts the filling contents of the at least two forms with a chain relationship based on a set model cluster. Each of the at least two forms with a chain relationship corresponds to a model in the set model cluster, and the chain relationship represents that the at least two forms have a filling order, and the filling content of the first form is determined based on the filling content of the second form, and the filling order corresponding to the first form is located before the filling order corresponding to the second form. The embodiment of the present invention solves the problem that the related art cannot predict the filling contents of multiple forms at one time, and can predict the filling contents of multiple forms with a chain relationship. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a schematic diagram of an implementation flow of a form content prediction method provided by an embodiment of the present invention;

[0042] Figure 2 It is a schematic diagram of the implementation flow of another form content prediction method provided by an embodiment of the present invention;

[0043] Figure 3 It is a schematic diagram of the implementation flow of another form content prediction method provided by an embodiment of the present invention;

[0044] Figure 4 It is a structural diagram of a form content prediction system provided by an application embodiment of the present invention;

[0045] Figure 5 It is a structural schematic diagram of a model module provided by an application embodiment of the present invention;

[0046] Figure 6 is a schematic diagram of a form content prediction device provided by an embodiment of the present invention;

[0047] Figure 7 It is a schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] In the e-commerce field, when users place orders, return goods, exchange goods, or apply for after-sales service at the terminal, a corresponding form information filling request will be generated in the system background. For example, when a user places an order, an order information filling request will be generated. Sometimes, a user's operation will generate multiple form information filling requests. For example, when a user applies for after-sales service, a courier information filling request and an after-sales order information filling request will be generated.

[0050] These forms are huge in number and have complicated contents. If they are all filled out manually, it would take a lot of effort. Related technologies use deep learning models to predict the contents of forms, thereby automatically filling out forms. However, deep learning models only support single-round interaction, that is, they can only predict the contents of one form at a time, and lack the ability to interact in multiple rounds, that is, the ability to predict the contents of multiple forms at a time. Especially when there are requirements for the order of filling out multiple forms, such as when the content of the next form must depend on the content of the form item filled in the previous step, the related technology cannot predict the contents of multiple forms.

[0051] In view of the shortcomings of the above-mentioned related technologies, the embodiments of the present invention provide a form content prediction method, which can at least solve the problem that the related technologies cannot predict the filled-in contents of multiple forms at one time. In order to illustrate the technical solution of the present invention, a specific embodiment is used for illustration.

[0052] Figure 1 1 is a schematic diagram of an implementation flow of a form content prediction method provided by an embodiment of the present invention. The execution subject of the form content prediction method may be an electronic device, and the electronic device may be a desktop computer, a laptop computer, a server, etc. Figure 1 , the form content prediction methods include:

[0053] S101, receiving a filling request; the filling request is used to request filling in a form.

[0054] Here, the filling request is generated based on the user's operation. For example, when a user applies for a return, a filling request for the return form will be generated; when a user applies for after-sales service, a filling request for the express delivery form information and a filling request for the after-sales form information will be generated.

[0055] In actual application, the user performs the return operation on the return page, fills in information such as the returned items, price, return reason and courier address. After this information is entered into the return page, the user clicks the return button on the return page. At this time, the system background will generate a request to fill out the return form.

[0056] S102, when the filling request is associated with at least two forms, determine whether the at least two forms include at least two forms with a chain relationship; the chain relationship indicates that the at least two forms have a filling sequence, and the filling content of the first form is determined based on the filling content of the second form; the filling sequence corresponding to the first form is located before the filling sequence corresponding to the second form.

[0057] In some cases, a filling request will be associated with at least two forms. For example, when a user applies for after-sales service, a filling request for express delivery order information and a filling request for after-sales order information will be generated, and the filling request is associated with the express delivery form and the after-sales form; when a user applies for a return, a filling request for express delivery order information and a filling request for return order information will be generated, and the filling request is associated with the express delivery form and the return form.

[0058] In the case where a filling request is associated with at least two forms, the at least two forms may include at least two forms with a chain relationship, and the chain relationship means that these forms have a filling order, and the filling content of the first form needs to be determined based on the filling content of the second form, wherein the filling order corresponding to the first form is before the filling order corresponding to the second form. Here, the first form and the second form are only used to illustrate the filling order of the forms, the first form refers to the form that needs to be filled in first, and the second form refers to the form that needs to be filled in after the first form is filled in. For example, when a user applies for a return, a filling request for express delivery information and a filling request for return delivery information will be generated, and the express delivery form and the return delivery form will be associated. Since the express delivery number needs to be filled in the return delivery form, the express delivery form needs to be filled in first, and then the return delivery form.

[0059] S103, when the at least two forms include at least two forms with a chain relationship, predict the filled-in contents of the at least two forms with a chain relationship based on a set model cluster; each form in the at least two forms with a chain relationship corresponds to a model in the set model cluster.

[0060] In a case where the at least two forms include at least two forms having a chain relationship, the filled-in contents of the at least two forms having the chain relationship are predicted based on the set model cluster.

[0061] When predicting the filled-in contents of the at least two forms having a chain relationship based on the set model cluster, the method includes:

[0062] The filled-in content of the first form is input into the first model corresponding to the second form in the set model cluster to obtain the filled-in content of the second form predicted by the corresponding first model.

[0063] For example, at least two forms with a chain relationship are Form 1, Form 2 and Form 3, and the order of filling in the forms is Form 1-Form 2-Form 3. The set model cluster includes Model 1, Model 2 and Model 3. First, the filling content of Form 1 is predicted by Model 1, and then the predicted filling content of Form 1 is input into Model 2 to predict the filling content of Form 2. Then, the filling content of Form 1 and Form 2 is input into Model 3 to predict the filling content of Form 3. Among them, the input of Model 1 is the information entered by the user. For example, if the user returns the goods, the input of Model 1 includes the return goods, return address and order number selected by the user. It should be understood that the above scheme has a sequential requirement in terms of the overall time, but there is no requirement in terms of the time interval, which improves the business scenarios of the e-commerce retail industry. For example, the filling of the order form is created on the first day, and the form filling of the waybill information is created again on the second day due to the delay of the courier.

[0064] In practical applications, the model in the set model cluster can be trained according to the filled-in content of historical forms. The filled-in content of historical forms is the filled-in content of forms with chain relationships. After multiple rounds of iterative training, the trained model is stored in the model information library, and the model is called from the model information library when it is needed.

[0065] In one embodiment, the method further comprises:

[0066] In the case where the filling request is associated with only one form, the filling content of the form is predicted based on the set second model; the set second model is used to predict the filling content of the form associated with the filling request associated with only one form.

[0067] When a filling request is associated with only one form, that is, only one form needs to be filled in, the filled content of the form is predicted based on the set second model. In practical applications, the filled content of historical forms can be used to train the set second model, where the filled content of historical forms is the filled content of the form associated with the filling request associated with only one form. After multiple rounds of iterative training, the trained model is stored in the model information library, and the model is called from the model information library when it is needed.

[0068] In one embodiment, the method further comprises:

[0069] In the case where the at least two forms do not include at least two forms with a chain relationship, the filled-in content of each of the at least two forms is predicted based on a set third model; the set third model is used to predict the filled-in content of the at least two forms that do not include the at least two forms with a chain relationship.

[0070] In the case where the filling request is associated with at least two forms, and the at least two forms do not include at least two forms with a chain relationship, the filling content of each of the at least two forms is predicted based on the set third model. In practical applications, the set second model can be trained using the filling content of historical forms, where the filling content of historical forms is the filling content that does not include at least two forms with a chain relationship. After multiple rounds of iterative training, the trained model is stored in the model information library, and the model is called from the model information library when the model is needed.

[0071] In the above scheme, the corresponding model is selected for prediction according to the form associated with the filling request. In the case where the filling request is associated with only one form, the filling content of the form is predicted based on the set second model. In the case where the filling request is associated with at least two forms, if the at least two forms include at least two forms with a chain relationship, the filling content of the at least two forms with a chain relationship is predicted based on the set model cluster. If the at least two forms do not include at least two forms with a chain relationship, the filling content of each form in the at least two forms is predicted based on the set third model.

[0072] If the filling request is associated with at least two forms, and the at least two forms include at least two forms with a chain relationship and other forms without a chain relationship, the filling contents of the at least two forms with a chain relationship are predicted based on the set model cluster, and the filling contents of the other forms are predicted based on the set second model or the set third model. If the number of other forms is 1, the prediction is made through the set second model; if the number of other forms is greater than or equal to 2, the prediction is made through the set third model.

[0073] The embodiment of the present invention receives a filling request, and in the case where the filling request is associated with at least two forms, determines whether the at least two forms include at least two forms with a chain relationship, and in the case where the at least two forms include at least two forms with a chain relationship, predicts the filling contents of the at least two forms with a chain relationship based on a set model cluster. Each of the at least two forms with a chain relationship corresponds to a model in the set model cluster, and the chain relationship represents that the at least two forms have a filling order, and the filling content of the first form is determined based on the filling content of the second form, and the filling order corresponding to the first form is located before the filling order corresponding to the second form. The embodiment of the present invention solves the problem that the related art cannot predict the filling contents of multiple forms at one time, and can predict the filling contents of multiple forms with a chain relationship.

[0074] refer to Figure 2 In one embodiment, the method further comprises:

[0075] S201, monitor at least one indicator corresponding to each model in at least one model.

[0076] The at least one model includes at least one of the following:

[0077] All models in the set model cluster;

[0078] The second model of setting;

[0079] The third model of the setting.

[0080] In one embodiment, the at least one indicator includes at least any one of the following:

[0081] Accuracy: The accuracy represents the ratio of the number of correct predictions made by the corresponding model within a set time period to the total number of predictions, that is,

[0082]

[0083] For example, if the model made 20 predictions in 1 hour, and 16 of them were accurate, the accuracy rate would be 0.8.

[0084] Precision: The precision represents the ratio of the number of times the corresponding model predicts the correct form content within a set time period to the number of times the model predicts the correct form category, that is,

[0085]

[0086] The form has a corresponding category. For example, when a user returns a product, the form category is "return". The number of times the form category is predicted here is the number of times the model correctly predicts the form category. For example, the model predicts 20 times in 1 hour, of which 18 times predict the form category correctly and 20 times predict the form content correctly, then the accuracy is 0.9.

[0087] Adoption rate: The adoption rate represents the proportion of the number of times the form content is filled in the form in the number of times the corresponding model predicts correctly within a set time period, that is,

[0088]

[0089] The actual number of adoptions refers to the number of times the content is actually filled in the form. For example, the model made 20 predictions in 1 hour, of which 16 were correct. Of these 16 times, 15 of the predicted form contents were filled in, and one correct form content was not filled in. The adoption rate is 0.9375.

[0090] The assistance rate represents the ratio of the number of filled-in forms that are correctly predicted by the corresponding model to the total number of forms, that is,

[0091]

[0092] Here, the actual number of adoptions refers to the number of forms whose contents are correctly predicted by the model and filled in. For example, if the model predicts the contents of 10 forms at a time, and the contents of 8 forms are predicted successfully, but only the contents of 6 forms are filled in, the assistance rate is 0.6.

[0093] S202: When the value of any one of the at least one indicator is less than a set value, update the corresponding model.

[0094] The purpose of setting the above indicators is to use targeted indicators to automatically update the model. For example, during festivals such as the Double 11 promotion and the 618 promotion, the order volume of certain categories will surge, and the model must improve its adaptability to perform timely model correction and self-update.

[0095] The above accuracy, precision, adoption rate and assistance rate can effectively measure the changes in data distribution in the e-commerce retail field. For example, after a new mobile phone is released, the number of inquiries related to this mobile phone surges, resulting in a surge in related forms. At this time, since the model has no bias towards this category, the adoption rate drops to the set threshold, prompting the model to be updated.

[0096] The above indicators are monitored, and when the above indicators are less than the set value, the model is updated. For example, the set value corresponding to the accuracy is 0.8, and when the accuracy is detected to be less than 0.8, the model is updated. As the number of model predictions increases, the accuracy index value gradually stabilizes.

[0097] The above indicators can also be calculated on a scheduled basis, for example, starting to calculate real-time indicators at 9 am every day.

[0098] In addition to updating the model based on the above indicators, the model can also be updated regularly, such as updating the model at 12:00 p.m. every day. Alternatively, the model can be updated according to data writing, such as updating the model after archiving the data every day.

[0099] The model includes two self-update strategies: hot update and cold update. When the model indicator is lower than the set value, it indicates that the data distribution has changed. This is very common in retail scenarios. At this time, in addition to providing external prediction capabilities, the model performs cold updates internally and self-updates while ensuring that it does not affect online services. During cold updates, the old model is used to provide prediction services. In addition, customer service can also automatically apply to update the model. At this time, the hot update of the model is started. This is to prevent the need to use data to update the model in a timely manner due to special reasons. For example, customer service or backend personnel add a fourth-level category item to all order information and request an update. At this time, the matching entry of data information is completed first, and then the model is hot updated. At this time, customer service can choose to call the existing model in the model library for prediction, or wait for the hot update of the model to be completed.

[0100] In actual application, assuming the auxiliary rate is around 80%, after statistics, minus the manual re-examination and other losses, it can be directly converted into productivity and save at least 60% of directly related personnel. Assuming there are 600 directly related personnel, it can save about 360 people. Taking a monthly salary of 8,000 as an example, it can save about 2.88 million yuan per month.

[0101] refer to Figure 3 In one embodiment, the method further comprises:

[0102] S301, determining a form corresponding to the filled-in content of the form in a setting database based on set form items; the filled-in content of the form includes the set form items; the setting database is used to store forms.

[0103] S302, filling in the form content into the corresponding form.

[0104] In related technologies, the storage method of form data mainly relies on manually set storage rules. One form corresponds to one storage rule, and there is no global clue between forms. If a new list item is added to the form, the storage structure of the database must be changed, which causes resource waste and model failure. Moreover, in the field of e-commerce, order tracking often has a certain logic. Related technologies lack this tracking capability, resulting in the inability to accurately summarize form items.

[0105] In view of the problem that the related art has poor form data summarization capability, the embodiment of the present invention summarizes form data based on set form items. The form is stored in a setting database, and the form includes set form items. By searching the form in the setting database through the set form items, the form position corresponding to the filled-in content of the form can be accurately located, thereby realizing accurate summarization of the form data.

[0106] In one embodiment, the set form item includes any one of the following:

[0107] Order ID;

[0108] Express delivery note identification;

[0109] User ID.

[0110] For example, when a user returns a product, the content of the form includes an order identifier, which can be an order number. The setting database is searched according to the order number to find the form corresponding to the order number in the setting database, so as to fill in the content of the form into the corresponding form. The embodiment of the present invention does not need to modify the storage rules of the form data according to the changes in the form, which not only makes the storage of the form data more convenient, but also lays the foundation for introducing new feature items into the model. At the same time, it can also improve the data processing speed and simplify the data construction process.

[0111] In the field of e-commerce, the set form items serve as clues throughout the entire process. The set form items may include order numbers, waybill numbers, and user IDs (Identity documents). The form content is inserted into the form in two ways: insertion without clue values ​​and insertion with clue values. The clue value is the value corresponding to the set form item. Insertion without clue values ​​is suitable for new data entry, while insertion with clue values ​​is suitable for scenarios such as data completion and new data items.

[0112] refer to Figure 4 , Figure 4 It is a structural diagram of a form content prediction system provided by an application embodiment of the present invention. The form content prediction system includes: a model module, a monitoring module and an interaction module.

[0113] The model module includes a model information library and a data information library. The model information library stores various models, such as the set model cluster, the set second model and the set third model in the above embodiment. The data information library is used to store forms, such as return forms, express forms and after-sales forms. The form data in the data information library can be used to update the model in the model information library.

[0114] The interaction module is used to receive user requests, which include filling requests, real-time viewing requests and update requests. For example, when a user applies for an update request, the interaction module sends the update request to the model update device in the model module, and the model update device updates the model in the model information library.

[0115] When users return goods, place orders, or perform after-sales operations, they send a filling request to the interactive module. The interactive module calls the model in the model information library based on the filling request to predict the form filling content and intelligently fills the form. If multiple rounds of filling are involved, it is necessary to determine whether forms with chain relationships are included.

[0116] The monitoring module is used to monitor various indicators of the model, including accuracy, precision, adoption rate and assistance rate. Users can send real-time viewing requests to the interactive module, and the interactive module displays the indicator values ​​to users.

[0117] Check Figure 5 , Figure 5 It is a structural diagram of a model module provided by an application embodiment of the present invention, and the model module includes an input layer, a model layer and an output layer.

[0118] The model input layer is used to receive information input by users. For example, when the input is product category or warehouse number, the model extracts digital discrete features; when the input is product price or product rating, the model extracts digital continuous features; when the input is order log or user message, the model extracts text features. The input layer performs data processing, which is used to determine whether the filling request is associated with at least two forms.

[0119] The model layer sets the model parameters and model calling method according to the data processing results and calling strategy. When filling in a request and only one form is associated, the single calling method is selected; when filling in a request and multiple forms are associated, it is determined whether the multiple forms have a chain relationship. If there is a chain relationship, the chain calling method is selected; if there is no chain relationship, the synchronous multi-calling method is selected. Each calling method corresponds to a model. For example, the single calling method selects the second model set, the synchronous multi-calling method selects the third model set, and the chain calling method selects the set model cluster.

[0120] The model layer outputs the prediction results of the model. After packaging in the output layer, single-fill output and multi-fill output are distinguished and returned to the user.

[0121] The embodiments of the present invention provide single call, chain call and synchronous multiple calls for different filling orders and filling contents of various categories of forms in the e-commerce field, effectively support multi-round interaction methods for form filling needs, solve the problem that related technologies cannot predict the filling contents of multiple forms at one time, and can predict the filling contents of multiple forms with a chain relationship.

[0122] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0123] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0124] It should be noted that the technical solutions described in the embodiments of the present invention can be arbitrarily combined without conflict.

[0125] In addition, in the embodiments of the present invention, "first", "second", etc. are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0126] refer to Figure 6 , Figure 6 is a schematic diagram of a form content prediction device provided by an embodiment of the present invention, such as Figure 6 As shown, the device includes: a receiving module, a determining module and a predicting module.

[0127] A receiving module, used to receive a filling request; the filling request is used to request to fill in a form;

[0128] A determination module, configured to determine, when the filling request is associated with at least two forms, whether the at least two forms include at least two forms having a chain relationship; the chain relationship indicates that the at least two forms have a filling sequence, and wherein the filling content of the first form is determined based on the filling content of the second form; and the filling sequence corresponding to the first form is before the filling sequence corresponding to the second form;

[0129] A prediction module is used to predict the filled-in contents of the at least two forms with a chain relationship based on a set model cluster when the at least two forms include at least two forms with a chain relationship; each form in the at least two forms corresponds to a model in the set model cluster.

[0130] In one embodiment, when the prediction module predicts the filled-in contents of the at least two forms having a chain relationship based on the set model cluster, it is used to:

[0131] The filled-in content of the first form is input into the first model corresponding to the second form in the set model cluster to obtain the filled-in content of the second form predicted by the corresponding first model.

[0132] In one embodiment, the prediction module is further configured to:

[0133] In the case where the filling request is associated with only one form, the filling content of the form is predicted based on the set second model; the set second model is used to predict the filling content of the form associated with the filling request associated with only one form.

[0134] In one embodiment, the prediction module is further configured to:

[0135] In the case where the at least two forms do not include at least two forms with a chain relationship, the filled-in content of each of the at least two forms is predicted based on a set third model; the set third model is used to predict the filled-in content of the at least two forms that do not include the at least two forms with a chain relationship.

[0136] In one embodiment, the device further comprises:

[0137] A monitoring module, used to monitor at least one indicator corresponding to each model in at least one model;

[0138] An updating module, configured to update the corresponding model when the value of any one of the at least one indicator is less than a set value;

[0139] The at least one model includes at least one of the following:

[0140] All models in the set model cluster;

[0141] The second model of setting;

[0142] The third model of the setting.

[0143] In one embodiment, the at least one indicator includes at least any one of the following:

[0144] Accuracy rate; the accuracy rate represents the ratio of the number of correct predictions made by the corresponding model within a set time period to the total number of predictions;

[0145] Precision rate; the precision rate represents the ratio of the number of times the corresponding model correctly predicts the form content to the number of times the model correctly predicts the form category within the set time period;

[0146] Adoption rate; the adoption rate represents the proportion of the number of times the form content is filled in the form among the number of times the corresponding model predicts correctly within the set time period;

[0147] Assistance rate; the assistance rate represents the ratio of the number of filled-in forms among the filled-in forms predicted correctly by the corresponding model to the total number of forms.

[0148] In one embodiment, the device further comprises:

[0149] A form determination module, used to determine the form corresponding to the filled-in content of the form in the setting database based on the set form items; the filled-in content of the form includes the set form items; the setting database is used to store forms;

[0150] The filling module is used to fill in the filling content of the form into the corresponding form.

[0151] In one embodiment, the set form item includes any one of the following:

[0152] Order ID;

[0153] Express delivery note identification;

[0154] User ID.

[0155] In actual application, the receiving module, determination module and prediction module can be implemented by a processor in an electronic device, such as a central processing unit (CPU), a digital signal processor (DSP), a microcontroller unit (MCU) or a programmable gate array (FPGA).

[0156] It should be noted that: the form content prediction device provided in the above embodiment only uses the division of the above modules as an example when performing form content prediction. In actual applications, the above processing can be assigned to different modules as needed, that is, the internal structure of the device is divided into different modules to complete all or part of the above-described processing. In addition, the form content prediction device provided in the above embodiment and the form content prediction method embodiment belong to the same concept, and the specific implementation process is detailed in the method embodiment, which will not be repeated here.

[0157] Based on the hardware implementation of the above program modules and in order to implement the method of the embodiment of the present application, the embodiment of the present application also provides an electronic device. Figure 7 Schematic diagram of the hardware structure of the electronic device of the present application embodiment. Figure 7 As shown, the electronic equipment includes:

[0158] Communication interface, which can exchange information with other devices such as network equipment;

[0159] The processor is connected to the communication interface to realize information exchange with other devices, and is used to execute the method provided by one or more technical solutions of the electronic device side when running the computer program. The computer program is stored in the memory.

[0160] Of course, in actual applications, the various components in the electronic device are coupled together through the bus system. It is understandable that the bus system is used to achieve connection and communication between these components. In addition to the data bus, the bus system also includes a power bus, a control bus, and a status signal bus. However, for the sake of clarity, Figure 7 In the specification, various buses are labeled as bus systems.

[0161] The memory in the embodiment of the present application is used to store various types of data to support the operation of the electronic device. Examples of such data include: any computer program used to operate on the electronic device.

[0162] It can be understood that the memory can be a volatile memory or a non-volatile memory, and can also include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a magnetic random access memory (FRAM), a flash memory, a magnetic surface memory, an optical disk, or a compact disc read-only memory (CD-ROM); the magnetic surface memory can be a disk memory or a tape memory. The volatile memory can be a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available, such as static random access memory (SRAM), synchronous static random access memory (SSRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), direct memory bus random access memory (DRRAM). The memory 130 described in the embodiments of the present application is intended to include but is not limited to these and any other suitable types of memory.

[0163] The method disclosed in the above embodiment of the present application can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or an instruction in the form of software. The above processor may be a general-purpose processor, a DSP, or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The processor can implement or execute the various methods, steps and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor may be a microprocessor or any conventional processor, etc. In combination with the steps of the method disclosed in the embodiment of the present application, it can be directly embodied as a hardware decoding processor to execute, or it can be executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a storage medium, which is located in a memory, and the processor reads the program in the memory and completes the steps of the above method in combination with its hardware.

[0164] Optionally, when the processor executes the program, the corresponding processes implemented by the electronic device in each method of the embodiments of the present application are implemented, which will not be described in detail here for the sake of brevity.

[0165] In an exemplary embodiment, the present application also provides a storage medium, namely a computer storage medium, specifically a computer-readable storage medium, for example, including a first memory storing a computer program, and the computer program can be executed by a processor of an electronic device to complete the steps of the aforementioned method. The computer-readable storage medium can be a memory such as FRAM, ROM, PROM, EPROM, EEPROM, Flash Memory, magnetic surface storage, optical disk, or CD-ROM.

[0166] In the several embodiments provided in the present application, it should be understood that the disclosed devices, electronic devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0167] The units described above 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 distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0168] In addition, all functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may be separately configured as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0169] A person of ordinary skill in the art can understand that: all or part of the steps of implementing the above-mentioned method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium, which, when executed, executes the steps of the above-mentioned method embodiment; and the aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, disks or optical disks.

[0170] Alternatively, if the above-mentioned integrated unit of the present application is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the embodiment of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0171] It should be noted that the technical solutions described in the embodiments of the present application can be combined arbitrarily without conflict.

[0172] In addition, in the examples of the present application, "first", "second", etc. are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0173] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A form content prediction method, characterized in that: The method comprises: Receive a filling request; the filling request is used to request filling in a form; In the case where the filling request is associated with at least two forms, determining whether the at least two forms include at least two forms having a chain relationship; the chain relationship indicates that the at least two forms have a filling sequence, and the filling content of the second form is determined based on the filling content of the first form; and the filling sequence corresponding to the first form is before the filling sequence corresponding to the second form; In the case where the at least two forms include at least two forms with a chain relationship, predicting the filled-in contents of the at least two forms with a chain relationship based on a set model cluster; each form in the at least two forms with a chain relationship corresponds to a model in the set model cluster; The predicting of the filled-in contents of the at least two forms having a chain relationship based on the set model cluster includes: The filled-in content of the first form is input into the first model corresponding to the second form in the set model cluster to obtain the filled-in content of the second form predicted by the corresponding first model.

2. The method according to claim 1, characterized in that The method further comprises: In the case where the filling request is associated with only one form, the filling content of the form is predicted based on the set second model; the set second model is used to predict the filling content of the form associated with the filling request associated with only one form.

3. The method according to claim 1, characterized in that The method further comprises: In the case where the at least two forms do not include at least two forms with a chain relationship, the filled-in content of each of the at least two forms is predicted based on a set third model; the set third model is used to predict the filled-in content of the at least two forms that do not include the at least two forms with a chain relationship.

4. The method according to any one of claims 1 to 3, characterized in that: The method further comprises: Monitoring at least one indicator corresponding to each model in at least one model; When the value of any one of the at least one indicator is less than a set value, updating the corresponding model; The at least one model includes at least one of the following: All models in the set model cluster; The second model of setting; The third model of the setting.

5. The method according to claim 4, characterized in that The at least one indicator includes at least one of the following: Accuracy rate; the accuracy rate represents the ratio of the number of correct predictions made by the corresponding model within a set time period to the total number of predictions; Precision rate; the precision rate represents the ratio of the number of times the corresponding model correctly predicts the form's filled-in content to the number of times the model correctly predicts the form's category within a set time period; Adoption rate; the adoption rate represents the proportion of the number of times the form content is filled in the form out of the number of times the corresponding model predicts correctly within a set time period; Assistance rate; the assistance rate represents the ratio of the number of filled-in forms among the filled-in forms predicted correctly by the corresponding model to the total number of forms.

6. The method according to any one of claims 1 to 3, characterized in that: The method further comprises: Determine a form corresponding to the filled-in content of the form in a setting database based on the set form items; the filled-in content of the form includes the set form items; the setting database is used to store forms; Fill in the form contents into the corresponding form.

7. The method according to claim 6, characterized in that The form items of the setting include any of the following: Order ID; Express delivery note identification; User ID.

8. A form content prediction device, characterized in that: include: A receiving module, used for receiving a filling request; The filling request is used to request filling in a form; A determination module, configured to determine, when the filling request is associated with at least two forms, whether the at least two forms include at least two forms having a chain relationship; the chain relationship indicates that the at least two forms have a filling sequence, and the filling content of the second form is determined based on the filling content of the first form; and the filling sequence corresponding to the first form is before the filling sequence corresponding to the second form; A prediction module, configured to predict the filled-in contents of the at least two forms having a chain relationship based on a set model cluster when the at least two forms include at least two forms having a chain relationship; each form in the at least two forms corresponds to a model in the set model cluster; The prediction module is used to input the filled content of the first form into the first model corresponding to the second form in the set model cluster, and obtain the filled content of the second form predicted by the corresponding first model.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the form content prediction method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the processor executes the form content prediction method according to any one of claims 1 to 7.

Images