Intelligent entry form method, electronic device and computer-readable storage medium

Abstract

The present application provides a method, electronic device, and computer-readable storage medium for intelligent form entry. The method of the present application mines a correlation model obtained by mining the correlation rules between fields in a specific scenario based on massive historical data of a specific scenario. During the form entry process, in response to an entry start operation for a first field in the form, the content information of a second field already entered in the form is obtained; the content information of the second field is input into the correlation model; the first recommended content of the first field is determined by the correlation model based on the content information of the second field and the correlation information between the second field and the first field; and intelligent entry of the first field is achieved based on the first recommended content of the first field. The method can perform correlation analysis based on the content information of the entered field through the correlation model, and intelligently recommend the content of the field that has not yet been entered, thereby achieving intelligent entry, making form entry more intelligent and greatly improving the efficiency of form entry.

Description

Technical Field

[0001] The present application relates to artificial intelligence technology, and in particular to a method for intelligently entering forms, an electronic device, and a computer-readable storage medium. Background Art

[0002] In the daily customs declaration process, document entry is a crucial and indispensable step for customs brokers. It also carries the heaviest workload and places the highest demands on timeliness. Improving form entry efficiency and reducing this workload is a long-term focus and investment area for customs declaration SaaS (Software as a Service). The customs declaration industry is a typical complex B2B application scenario, and customs declaration orders are also typical complex entry forms.

[0003] In application scenarios involving various form entry, such as customs declaration, current configurable form tools mostly focus on how to build applications that meet service flow requirements. However, the form entry process is not intelligent enough and the entry efficiency is low. Summary of the Invention

[0004] The present application provides a method for intelligently entering a form, an electronic device, and a computer-readable storage medium.

[0005] In one aspect, the present application provides a method for intelligently entering a form, comprising:

[0006] In response to an entry start operation on a first field in a form, obtaining content information of a second field entered in the form;

[0007] inputting the content information of the second field into an association model, and determining, through the association model, first recommended content of the first field based on the content information of the second field and association information between the second field and the first field;

[0008] Intelligent entry of the first field is achieved according to the first recommended content of the first field.

[0009] In another aspect, the present application provides an electronic device, comprising:

[0010] a processor, and a memory communicatively connected to the processor;

[0011] The memory stores computer-executable instructions;

[0012] The processor executes the computer-executable instructions stored in the memory to implement the above-mentioned method for intelligently entering a form.

[0013] On the other hand, the present application provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the above-mentioned method of intelligent entry form.

[0014] The method, electronic device and computer-readable storage medium for intelligent form entry provided by the present application use massive historical data based on specific scenarios to mine the association rules between fields in specific scenarios to obtain an association model. During the form entry process, in response to the entry start operation of the first field in the form, the content information of the second field that has been entered in the form is obtained; the content information of the second field is input into the association model, and the first recommended content of the first field is determined through the association model based on the content information of the second field and the association information between the second field and the first field; based on the first recommended content of the first field, intelligent entry of the first field is achieved, and association analysis can be performed through the association model based on the content information of the entered field, and the content of the fields that have not yet been entered can be intelligently recommended to achieve intelligent entry, making form entry more intelligent and greatly improving the efficiency of form entry. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0016] Figure 1 A flowchart of a method for intelligent form entry provided in one embodiment of the present application;

[0017] Figure 2 A flowchart of a method for intelligent entry of a form provided in another embodiment of the present application;

[0018] Figure 3 A schematic diagram of the framework of the intelligent recognition technology provided in one embodiment of the present application;

[0019] Figure 4 A schematic diagram of the automatic input technical framework provided in one embodiment of the present application;

[0020] Figure 5 A schematic diagram of a technical architecture for collecting data based on tracking technology provided in one embodiment of the present application;

[0021] Figure 6 A schematic structural diagram of a device for intelligently entering a form provided in another exemplary embodiment of the present application;

[0022] Figure 7 A schematic structural diagram of an electronic device provided in an exemplary embodiment of the present application.

[0023] The above drawings illustrate specific embodiments of the present application, which will be described in more detail below. These drawings and the textual description are not intended to limit the scope of the present application in any way, but rather to illustrate the concepts of the present application to those skilled in the art by reference to specific embodiments. DETAILED DESCRIPTION

[0024] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

[0025] The terms "first," "second," "third," etc., used in this application are for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the technical features indicated. In the description of the following embodiments, "multiple" means more than two, unless otherwise specifically defined.

[0026] In the daily customs declaration process, document entry is a crucial and indispensable step for customs brokers. It also carries the heaviest workload and demands the highest timeliness. Improving efficiency and reducing workload for document entry personnel is an area of ​​long-term focus and investment for customs declaration SaaS software. The customs declaration industry exemplifies complex B2B demand, and customs declaration orders are typically complex entry forms. This has become a bottleneck in efficiency, driving customs declaration SaaS software to advance further and faster in terms of intelligent form entry and improved efficiency.

[0027] The customs declaration industry has a vast amount of offline data, resulting in heavy data entry workloads and a massive amount of accumulated data, yet this data remains largely unutilized. By mining historical patterns, we can effectively empower new order entry and significantly improve efficiency. The intricate relationships between fields also help order clerks complete the entire customs declaration process more efficiently and intelligently. Customs declaration SaaS, through long-term practical experience, has accumulated extensive experience in intelligent data entry for complex order forms, placing it at the forefront. This expertise is evident in the platform's capabilities, developed based on specific scenarios.

[0028] In practice, form entry is used in numerous scenarios beyond customs declaration. For these and other form entry scenarios, current industry-leading products focus on form construction. Configurable form tools mostly focus on building server-side process applications to meet specific requirements. However, the form entry process lacks intelligence, resulting in low efficiency. Regarding how to make forms more intelligent in specific scenarios, no platform-based service products for order forms have yet been found that can be integrated and deployed by form application providers.

[0029] The method for intelligent form entry provided in this application can, based on massive historical data of specific scenarios, mine the association patterns between fields in specific scenarios and obtain an association model. During the form entry process, based on the content information of the entered fields, association analysis is performed through the association model, and the content of the fields that have not yet been entered is intelligently recommended to achieve intelligent entry, making form entry more intelligent and greatly improving the efficiency of form entry.

[0030] The method of intelligent form entry provided in this application brings together a full set of closed-loop intelligent form solutions, including intelligent import recognition, operation number tracking statistics (quantifiable), field association recommendation, intelligent recommendation, and automatic optimization based on recommendation adoption. It truly forms the concept of intelligent order recording, effectively improves the efficiency of order recorders, saves time and energy, reduces workload, and brings good economic benefits and humane care.

[0031] In addition, the intelligent form entry method provided in this application can be applied to more application fields, allowing more order form scenarios to enter the "intelligent" era and play a greater role. Whether it is a one-stop order, a credit guarantee order, or other complex Class B order form scenarios of the group, it has potential application value.

[0032] The following specific embodiments describe in detail the technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0033] Figure 1 Flowchart of the method for intelligent form entry provided in one embodiment of the present application. The method for intelligent form entry provided in this embodiment can be specifically applied to an electronic device that implements the form entry function, which can be a terminal device, a server, etc. In other embodiments, the electronic device can also be implemented using other devices, which is not specifically limited in this embodiment.

[0034] like Figure 1 As shown, the specific steps of this method are as follows:

[0035] Step S101: In response to an entry start operation on a first field in a form, content information of a second field entered in the form is obtained.

[0036] The first field may be any field in the form, and the second field may be any field entered in the form.

[0037] The entry start operation is used to trigger the intelligent entry of the first field, and may be an operation in which the entry person clicks to select the input area of ​​the first field when entering the first field.

[0038] In this embodiment, when entering any field (first field) in the form, the content information of other fields (second field) already entered in the current form is obtained, and the popular content information of the first field is intelligently associated based on the content information of the second field as the first recommended content of the first field.

[0039] Step S102: Input the content information of the second field into the association model, and determine the first recommended content of the first field according to the content information of the second field and the association information between the second field and the first field through the association model.

[0040] The association model is trained based on the current scenario and includes candidate content information for each field in each form type, as well as association information between fields. Each association information includes a mapping rule between at least one input field and an output field, which is used to determine candidate content information for the output field that meets the mapping rule based on the content information of the at least one input field.

[0041] After obtaining the content information of the entered second field, the content information of the entered second field is input into the association model. If the association information contained in the association model includes a mapping rule for mapping at least one second field to the first field, one or more candidate content information of the first field can be determined based on the content information of the at least one second field and the mapping rule, and the one or more candidate content information of the first field can be used as the first recommended content of the first field.

[0042] Step S103: Implement intelligent entry of the first field according to the first recommended content of the first field.

[0043] Optionally, after determining the first recommended content of the first field, the first recommended content may be output first, the user may select one of the first recommended contents, and the first recommended content selected by the user may be automatically entered to implement intelligent entry of the first field.

[0044] Optionally, after determining the first recommended content for the first field, the first recommended content with the highest priority can be automatically entered based on the priority order of the first recommended content, thereby achieving intelligent entry of the first field. The priority order can be sorted based on the popularity of the first recommended content, the number of times it has been adopted, etc.

[0045] In an embodiment of the present application, based on massive historical data based on specific scenarios, an association model is obtained by mining the association rules between fields in specific scenarios. During the form entry process, in response to the entry start operation of the first field in the form, the content information of the second field that has been entered in the form is obtained; the content information of the second field is input into the association model, and the first recommended content of the first field is determined through the association model based on the content information of the second field and the association information between the second field and the first field; based on the first recommended content of the first field, intelligent entry of the first field is realized, and based on the content information of the entered field, association analysis can be performed through the association model, and the content of the fields that have not yet been entered can be intelligently recommended to realize intelligent entry, making form entry more intelligent and greatly improving the efficiency of form entry.

[0046] Figure 2 This is a flow chart of a method for intelligent entry form provided by another embodiment of the present application. Figure 1 On the basis of the corresponding embodiment, in this embodiment, intelligent entry of the first field is implemented according to the first recommended content of the first field, including: if there is one and only one first recommended content of the first field, the first recommended content of the first field is automatically entered into the entry area of ​​the first field; if there are multiple first recommended contents of the first field, the first recommended content of the first field is output; in response to the selection operation of any first recommended content, the selected first recommended content is automatically entered into the entry box of the first field.

[0047] like Figure 2 As shown, the specific steps of this method are as follows:

[0048] Step S201: In response to an entry start operation on a first field in a form, content information of a second field entered in the form is obtained.

[0049] The first field may be any field in the form, and the second field may be any field entered in the form.

[0050] The entry start operation is used to trigger the intelligent entry of the first field, and may be an operation in which the entry person clicks to select the input area of ​​the first field when entering the first field.

[0051] In this embodiment, when entering any field (first field) in the form, the content information of other fields (second field) already entered in the current form is obtained, and the popular content information of the first field is intelligently associated based on the content information of the second field as the first recommended content of the first field.

[0052] Step S202: Input the content information of the second field into the association model, and determine the first recommended content of the first field according to the content information of the second field and the association information between the second field and the first field through the association model.

[0053] The association model is trained based on historical data for the current scenario and includes candidate content information for each field in each form type, as well as association information between fields. Each association information includes a mapping rule between at least one input field and an output field, which is used to determine candidate content information for the output field that satisfies the mapping rule based on the content information of the at least one input field.

[0054] After obtaining the content information of the entered second field, the content information of the entered second field is input into the association model. If the association information contained in the association model includes a mapping rule for mapping at least one second field to the first field, one or more candidate content information of the first field can be determined based on the content information of the at least one second field and the mapping rule, and the one or more candidate content information of the first field can be used as the first recommended content of the first field.

[0055] Optionally, the trained association model further includes: a field record count threshold and / or a field diversity threshold.

[0056] The field record count threshold is used to filter out fields with too little historical content information. Considering that too little historical data may lead to insufficient recommendation accuracy, fields with historical content information less than the field record count threshold are not recommended.

[0057] The field diversity threshold is used to filter out fields with insufficient historical content information diversity. Considering that insufficient historical data diversity may lead to inaccurate recommendations, fields with historical content information diversity parameters less than the field diversity threshold are not recommended.

[0058] The field record count threshold and field diversity threshold can be set and adjusted based on historical data in actual application scenarios and are not specifically limited here.

[0059] For example, taking the trained association model including the field record count threshold and the field diversity threshold as an example, this step can be implemented in the following manner:

[0060] The content information of the second field is input into the association model. Through the association model, if it is determined that the number of records of the historical content information of the first field is greater than or equal to the field record count threshold, and the diversity parameter of the first field is greater than or equal to the field diversity threshold, then the first recommended content of the first field is determined based on the content information of the second field and the association information between the second field and the first field.

[0061] If it is determined that the number of records of historical content information of the first field is less than the field record count threshold, or the diversity parameter of the first field is less than the field diversity threshold, there is no need to determine the first recommended content of the first field.

[0062] In this embodiment, the association model is trained based on historical data from the current scenario. Exemplarily, popular content information for each field included in each form type is obtained; based on the popular content information for each field included in each form type, the association model is trained, and a field record count threshold and a field diversity threshold are determined. The trained association model includes candidate content information for each field included in each form type, and at least one piece of association information, wherein each piece of association information includes a mapping rule between at least one input field and an output field, and is used to determine candidate content information for the output field that satisfies the mapping rule based on the content information of at least one input field.

[0063] The popular content information for any field can be content information that has appeared in the historical content information for that field and has a high popularity. For example, content information that has appeared in the historical content information for that field more than a specified threshold number of times, or that accounts for more than a specified percentage of content information. The specified threshold number of times can be set and adjusted based on actual application scenarios and is not specifically limited here.

[0064] Furthermore, to obtain the popular content information of each field contained in each form type, you can use the following method to achieve it:

[0065] Obtain imported historical form files; identify and process each historical form file to determine the form type corresponding to the historical form file and the historical content information of each field in the historical form file; aggregate the historical content information of the same field according to the field type to determine the historical content information that has appeared in each field and the frequency of occurrence of each historical content information; determine the popular content information of each field based on the frequency of occurrence of each historical content information of each field.

[0066] First, by identifying and processing each historical form file, we determine the corresponding form type and the historical content information of each field in the historical form file. This process breaks down the massive amount of form data accumulated in the current domain into form field granularity, and determines the historical content information of each field. Then, by field type, we aggregate the occurrence frequency information of each field's historical content information according to different conditions. Based on the occurrence frequency information of each field's historical content information, we summarize and generate the popular content information for each field. The popular content information of a field can be the frequently occurring historical content information within the field's historical content information.

[0067] For example, in actual applications, the content information of the same field may be different when applied to different demand logics. Therefore, the conditions for field aggregation can be set according to the demand logic, and the fields can be aggregated based on different conditions to obtain the high-frequency historical content information of the field under different demand logics. For example, different products have different values ​​for ingredient content. The possible values ​​of the ingredient content field for the same product code and product name can be aggregated based on the product code and product name as the aggregation basis to obtain the possible values ​​of the ingredient content of different products.

[0068] In this embodiment, the process of decomposing form data into form field granularity and aggregating each field according to conditions can be implemented by calling the corresponding service interface (such as FaaS function). The specific implementation method is not specifically limited here.

[0069] For example, the following is combined Figure 3 The following is an example of the process of obtaining imported historical form files through intelligent recognition technology, identifying any form file, determining the form type corresponding to the historical form file, and the historical content information of each field in the historical form file:

[0070] 1) User imports historical form files

[0071] Users can import historical form files of the current application scenario in batches or one by one. The imported historical form files can be any of the following types of files: Word, Excel, Image, Text, PDF, PNG, compressed files (such as zip, rar, etc.).

[0072] 2) Generate file preprocessing tasks

[0073] Based on the imported historical form files, a file preprocessing task corresponding to the historical form files is generated to provide pre-support for subsequent domain sorting tasks and domain extraction tasks. When the file preprocessing task is executed, the historical form files are preprocessed.

[0074] 3) File preprocessing

[0075] The preprocessing of historical form files includes: file format conversion, OCR (Optical Character Recognition) call and OCR marking and other recognition preparation work.

[0076] Exemplarily, the preprocessing of historical form files includes: converting the historical form files into files of a specified format (such as PDF, etc.); performing OCR calls on the obtained files of the specified format, extracting the text content and coordinate information in the files of the specified format, performing word segmentation processing based on the text content and coordinate information, and obtaining the phrases contained in the files of the specified format; performing OCR marking on the phrases to obtain preprocessed OCR marked text.

[0077] 4) Generate domain sorting tasks and / or domain extraction tasks

[0078] Among them, the domain sorting task is responsible for classifying the form type to which the OCR marked text belongs based on the OCR marked text of historical form files and all form types in the current domain scenario.

[0079] For example, form types in the customs declaration scenario may include customs declaration form, contract, packing list, invoice, power of attorney, etc.

[0080] The domain extraction task is responsible for extracting the domain model entities in the current domain scenario based on the file information or OCR-marked text of historical form files. When executing the domain extraction task, multiple different types of channels can be used, and different types of channels are used to process different types of data.

[0081] Exemplarily, at least one of the following channels may be included: an Excel recognition channel, an OCR recognition channel, an NLP recognition channel, etc.

[0082] For example, for Excel files, the domain extraction task can be implemented through the Excel recognition channel; for OCR-marked text, the domain extraction task can be implemented through the OCR recognition channel.

[0083] 5) Field sorting and field extraction

[0084] During the field sorting process, the form type of the historical form file is identified based on the OCR marked text of the historical form file; the form data of various form types are sorted out by merging and classifying based on the form type; and the historical form files of the same form type are compressed into files in the corresponding format of the form type in the current field scenario.

[0085] During the domain extraction process, an extraction channel is selected based on the file type of the historical form file. The domain model is extracted using the selected extraction channel. After processing through pre-set post-rules, domain model entities are generated, thereby determining the historical content information of each field in the historical form file. Post-rules can be set and adjusted based on the needs of actual domain scenarios and are not specifically limited here.

[0086] In addition, the definition of the domain model can be quickly replicated to scenarios in other domains. Among them, the post-rules can be set and adjusted according to the needs of the actual domain scenario, and are not specifically limited here.

[0087] In an optional implementation of this embodiment, an intelligent interactive feedback mechanism can be added to update the priority recommendation order of the candidate content information of the field based on the recommendation feedback results of the recommended content of the field in the form, so that subsequent recommendations are more accurate. Among them, the recommendation feedback result is adopted or not adopted. If the recommendation feedback result is adopted, it means that the recommended content recommended for the field is adopted by the user as the correct content information of the field. If the recommendation feedback result is not adopted, it means that the recommended content recommended for the field is not adopted by the user. It may be that the recommended content is automatically entered by the user but is stored after the user modifies it, or the recommended content is not selected by the user.

[0088] For example, when the form is submitted and stored, the recommendation feedback results of the recommended content of each field in the form are recorded, and the number of times each candidate content information of each field is adopted is counted.

[0089] Specifically, in response to a form submission operation, a recommendation feedback result of each recommended content in each field of the form is determined; and according to the recommendation feedback result of each recommended content, the number of times each candidate content information in each field is adopted is determined.

[0090] After determining the number of times each candidate content information of each field is adopted, the priority recommendation order of the candidate content information of each field may be updated according to the number of times each candidate content information of each field is adopted, so that subsequent recommendations are more accurate.

[0091] After determining the first recommended content of the first field, through steps S203-S206, intelligent entry of the first field is implemented according to the first recommended content of the first field.

[0092] Step S203: Determine whether there are multiple first recommended contents in the first field.

[0093] In this embodiment, the association model includes multiple pieces of association information, each of which includes a mapping rule between at least one input field and an output field, for determining candidate content information of the output field that meets the mapping rule based on the content information of the at least one input field.

[0094] In step S202, the entered content information of the second field may hit multiple mapping rules from the second field to the first field. Each mapping rule can generate at least one first recommended content of the first field, so there may be multiple first recommended contents of the first field.

[0095] In this step, if it is determined that there is only one first recommended content in the first field, step S204 is executed to automatically enter the first recommended content.

[0096] If there are multiple first recommended contents in the first field, steps S205 - S206 are executed, the user selects one first recommended content, and the selected first recommended content is automatically entered.

[0097] Step S204: If there is only one first recommended content in the first field, the first recommended content in the first field is automatically entered into the entry area of ​​the first field.

[0098] If it is determined that there is only one first recommended content in the first field, the first recommended content may be directly and automatically entered.

[0099] Optionally, if it is determined that there is only one first recommended content in the first field, the first recommended content may be output first, and then the first recommended content may be automatically entered in response to the user's confirmation operation.

[0100] Step S205: If there are multiple first recommended contents of the first field, output the first recommended contents of the first field.

[0101] In this step, if there are multiple first recommended contents in the first field, the first recommended contents in the first field may be output first, allowing the user to select one of the first recommended contents, and then the first recommended content selected by the user may be automatically entered.

[0102] Optionally, if there are multiple first recommended contents in the first field, some of the first recommended contents with higher priority can be selected for output according to the priority recommendation order of the first recommended contents, so that the user can select one of the first recommended contents by himself, and the first recommended content selected by the user is automatically entered.

[0103] Step S206 : In response to a selection operation on any first recommended content, the selected first recommended content is automatically entered into the entry box of the first field.

[0104] Optionally, if there are multiple first recommended contents in the first field, the first recommended contents may be sorted according to the number of hits of the mapping rules corresponding to the respective first recommended contents, and the first recommended content with the highest sorting order may be automatically entered in the entry area of ​​the first field.

[0105] Optionally, if there are multiple first recommended contents in the first field, the first recommended content with the highest priority can be automatically entered in the entry area of ​​the first field according to the priority order of each first recommended content.

[0106] In an optional implementation, while the user is manually entering the content of the first field, the input content that has been entered can be matched with the candidate content information of the first field. Based on the degree of matching between the input content and the candidate content information of the first field, the candidate content information with the highest degree of matching with the input content can be intelligently recommended to the user, and automatic entry can be achieved.

[0107] Optionally, after determining the first recommended content of the first field in step S202, if the user performs an entry operation in the first field, the first recommended content is further screened based on the matching degree between the entered input content and the first recommended content to recommend a second recommended content with higher accuracy. The second recommended content is output for the user to select and automatically enter, which can narrow the recommendation scope and improve the accuracy of the recommendation.

[0108] For example, after determining the first recommended content of the first field in step S202, the following steps S207 to S209 may be further included:

[0109] Step S207 : In response to the input operation on the first field in the form, according to the input content of the first field, the input content is matched with the first recommended content of the first field, and second recommended content matching the input content is determined.

[0110] Optionally, a similarity between the input content and the first recommended content in the first field can be calculated as a matching degree between the input content and the first recommended content, and the second recommended content includes the first recommended content whose matching degree exceeds a matching threshold. The matching threshold can be set and adjusted based on actual application scenarios and is not specifically limited here.

[0111] Optionally, the input content may be matched with the first recommended content of the first field to determine whether the first recommended content of the first field contains the input content, and the first recommended content containing the input content may be used as the second recommended content.

[0112] Step S208: Output second recommended content that matches the input content.

[0113] For example, the second recommended content can be displayed in a list format in a first designated area near the input area of ​​the first field, or in a second designated area on the right side of the current page, etc., to facilitate user browsing and selection. The specific location of the display area for the second recommended content is not specifically limited here.

[0114] Step S209: In response to the selection operation of any second recommended content, the selected second recommended content is automatically entered into the entry box of the first field.

[0115] In this step, the selected second recommended content is directly rendered into the input box of the first field without re-rendering the entire input page.

[0116] Optionally, a lightweight and flexible state manager called nice-core, designed for high-performance applications, can be implemented based on the react framework and event-driven programming. The state manager defines and processes the linkage relationships between fields. Based on the input content, the default content information or candidate content information of the linked fields is automatically entered, enabling intelligent recommendation and automatic filling capabilities for fields with millisecond-level response.

[0117] For example, based on Figure 4 The technical framework for automatic entry shown is that when a click or input operation on the entry box of the first field on the page triggers a corresponding event, the rule module performs rule routing according to the rules defined by the state manager. When intelligent recommendation is required, the back-end interface is called to perform intelligent recommendation based on the association model and intelligent recommendation based on input content matching, and multiple recommended contents are obtained. The recommended contents are associated and output to the corresponding location through the state manager (if a recommended content is intelligently recommended, the recommended content is the default recommended content). When a click operation on any of the output recommended contents on the page triggers a corresponding event, rule routing is performed according to the rules defined by the state manager. No intelligent recommendation is required. The selected recommended content is automatically rendered to the entry area corresponding to the first field through the state manager to achieve automatic entry.

[0118] In an optional implementation, before step S201, an existing form file may be imported, the content information of the fields contained in the form file may be intelligently identified, and the content information identified in the form file may be automatically entered.

[0119] Specifically, before step S201, the following steps may also be included:

[0120] Step S200: In response to the import operation of the form file of the form, the content information of the fields included in the form file is identified and automatically entered.

[0121] The step S200 can be implemented in the following manner:

[0122] In response to an import operation on a form file of a form, the imported form file is obtained, and the form content information contained in the form file is inconsistent with the structure of the specified template of the form; the form file is identified and processed to determine the text content of the form file and the form type of the form; the text content of the form file is structured according to the specified template corresponding to the form type of the form, and content information of at least one field of the form file is determined; the content information of the at least one field is automatically entered into the corresponding entry area.

[0123] The form file is any of the following file types: Word, Excel, Image, Text, or PDF.

[0124] The designated template corresponds to the form type. The designated template defines the structural information such as fields and field attributes contained in the form of the corresponding form type. Different form types correspond to different designated templates.

[0125] The imported form file can be any type of irregular file. By intelligently identifying and processing the irregular file, the fields and content information contained in the form file can be extracted, so that the content information of the identified field can be automatically entered into the corresponding entry area. There is no need for users to manually copy existing form files, making form entry more intelligent, greatly improving the efficiency of form entry, and reducing the workload of manual entry.

[0126] In actual applications, due to the individual differences and subjectivity of timeliness, timeliness is not a good quantitative indicator of the operational load and experience of form entry. In an optional implementation of this embodiment, the number of operations entered into each field can also be collected through the embedding technology. The number of operations entered into a field refers to the number of times the user taps the mouse and / or keyboard during the process of entering the field. Using the number of field entry operations as an indicator to measure the workload and operational load of form entry can clearly quantify the workload and operational load of form entry. If the number of operations entered into a field is reduced, it means that the frequency of user operations has been reduced, and the operational load of form entry has been reduced.

[0127] Optionally, through the tracking technology, the number of operations entered into each field of each form is collected. By summarizing the number of operations entered into each field of each form in multiple form entries in the current field scenario, the efficiency value of each field is calculated based on massive data (for example, it can be the average value of the number of entered operations, etc.), which can clearly quantify the workload and operational load of form entry.

[0128] For example, the technical architecture for collecting data based on tracking technology is as follows: Figure 5As shown, based on the tracking technology, the operation behavior data is collected by monitoring the DOM event information corresponding to each field generated during the form entry process, and the operation behavior data is filtered and reported according to the configured operation rules, and the operation behavior data is stored; statistics are performed based on the operation behavior data to determine the number of operations entered in each field, and a visual display can also be performed so that relevant personnel can understand the capabilities of smart form entry by viewing the information on the number of operations and perform subsequent optimization processing.

[0129] The operation rules can be configured according to the actual application scenario to filter the reported data. For example, the operation rules can be used to set the field set to be reported, and the operation behavior data related to the fields in the field set can be filtered and reported.

[0130] The intelligent form entry method provided in this embodiment enables a SaaS software platform for form entry, offering extremely convenient access. Historical data is imported into model training in a standard format, and forms integrate intelligent capabilities through JAR package components and API service calls, making overall packaging easy. Customs declaration SaaS, through long-term practical experience, has accumulated extensive experience in intelligent entry of complex forms and is a leader in this field. This area is also highly specialized, with platform capabilities developed based on demanding scenarios.

[0131] The intelligent form entry method provided in this embodiment can also be implemented by hard code, but the development cost is very huge.

[0132] The method for intelligent form entry provided by the embodiment of the present application has the ability to preliminarily structure offline irregular form data through intelligent recognition; the more entries are made, the more accurate the recommended content for the fields to be entered will be, and it can be entered automatically; it can respond in real time, will not slow down the entry efficiency, but can bring about a significant improvement in the efficiency of form entry; the recommended content for intelligent recommendation is not unique, and users can quickly select and automatically enter it, which is more intelligent and humane; it can quickly integrate intelligent form entry solutions in specific field scenarios at a relatively low cost, realize the intelligent entry of complex forms in specific fields, and realize a significant improvement in form entry experience and efficiency. By injecting historical data in different fields, it can be applied to the intelligent entry of forms in field scenarios, improving the intelligence and efficiency of form entry.

[0133] Figure 6 This is a structural diagram of a device for intelligently entering a form provided by another exemplary embodiment of the present application. The device for intelligently entering a form provided by the embodiment of the present application can execute the processing flow provided by the method embodiment of the intelligently entering a form. Figure 6As shown, the device 60 for intelligently entering a form includes: an entered information acquisition module 601 , an associated recommendation module 602 and an intelligent entry module 603 .

[0134] Specifically, the entered information acquisition module 601 is configured to acquire content information of a second field entered in the form in response to an entry start operation on a first field in the form.

[0135] The association recommendation module 602 is configured to input the content information of the second field into the association model, and determine the first recommended content of the first field according to the content information of the second field and the association information between the second field and the first field through the association model.

[0136] The intelligent entry module 603 is used to implement intelligent entry of the first field according to the first recommended content of the first field.

[0137] Optionally, the intelligent entry module is further used to:

[0138] If there is only one first recommended content in the first field, the first recommended content of the first field will be automatically entered into the entry area of ​​the first field; if there are multiple first recommended contents in the first field, the first recommended content of the first field will be output; in response to the selection operation of any first recommended content, the selected first recommended content will be automatically entered into the entry box of the first field.

[0139] Optionally, the device for intelligently entering a form may further include:

[0140] The second recommended module is for:

[0141] In response to an entry operation on the first field in the form, the input content is matched with the first recommended content of the first field based on the input content of the first field, and a second recommended content matching the input content is determined; the second recommended content matching the input content is output; in response to a selection operation on any second recommended content, the selected second recommended content is automatically entered into the entry box of the first field.

[0142] Optionally, the device for intelligently entering a form may further include:

[0143] Intelligent identification module for:

[0144] In response to an import operation on a form file of a form, the imported form file is obtained, and the form content information contained in the form file is inconsistent with the structure of the specified template of the form; the form file is identified and processed to determine the text content of the form file and the form type of the form; the text content of the form file is structured according to the specified template corresponding to the form type of the form, and content information of at least one field of the form file is determined; the content information of the at least one field is automatically entered into the corresponding entry area.

[0145] Optionally, the association recommendation module is further configured to:

[0146] The content information of the second field is input into the association model. If it is determined through the association model that the number of records of the historical content information of the first field is greater than or equal to the field record count threshold, and the diversity parameter of the first field is greater than or equal to the field diversity threshold, then the first recommended content of the first field is determined based on the content information of the second field and the association information between the second field and the first field.

[0147] Optionally, the device for intelligently entering a form may further include:

[0148] Model training module, used for:

[0149] Obtain popular content information for each field contained in each form type; train an association model based on the popular content information for each field contained in each form type, and determine a field record count threshold and a field diversity threshold, wherein the trained association model includes candidate content information for each field contained in each form type, and at least one piece of association information, wherein each piece of association information includes a mapping rule between at least one input field and an output field, and is used to determine candidate content information of the output field that meets the mapping rule based on the content information of at least one input field.

[0150] Optionally, the device for intelligently entering a form may further include:

[0151] Interactive feedback module, used to:

[0152] Obtain imported historical form files; identify and process each historical form file to determine the form type corresponding to the historical form file and the historical content information of each field in the historical form file; aggregate the historical content information of the same field according to the field type to determine the historical content information that has appeared in each field and the frequency of occurrence of each historical content information; determine the popular content information of each field based on the frequency of occurrence of each historical content information of each field.

[0153] Optionally, the interactive feedback module is further configured to:

[0154] In response to the submission operation of the form, the recommendation feedback result of each recommended content in each field of the form is determined, and the recommendation feedback result is adopted or not adopted; based on the recommendation feedback result of each recommended content, the number of times each candidate content information of each field is adopted is determined; based on the number of times each candidate content information of each field is adopted, the priority recommendation order of the candidate content information of each field is updated.

[0155] The device provided in the embodiments of the present application can be specifically used to execute the method embodiments provided in any of the above method embodiments, and the specific functions and effects will not be repeated here.

[0156] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an exemplary embodiment of the present application. Figure 7 As shown, the electronic device 70 includes: a processor 701 , a memory 702 , and a computer program stored in the memory 702 and executable on the processor 701 .

[0157] Among them, when the processor 701 runs the computer program, it implements the method of intelligent entry form provided by any of the above method embodiments. The specific functions and technical effects that can be achieved are not repeated here.

[0158] An embodiment of the present application also provides a computer-readable storage medium, which stores computer-executable instructions. When the computer-executable instructions are executed by a processor, they are used to implement the method of intelligent entry form provided by any of the above method embodiments.

[0159] An embodiment of the present application also provides a computer program product, which includes: a computer program, the computer program is stored in a readable storage medium, at least one processor of the picking device can read the computer program from the readable storage medium, and at least one processor executes the computer program so that the picking device executes the method of intelligent form entry provided by any of the above method embodiments.

[0160] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0161] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0162] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0163] The above-mentioned integrated unit implemented in the form of a software functional unit can be stored in a computer-readable storage medium. The above-mentioned software functional unit is stored in a storage medium and includes a number of instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) or a processor to execute some steps of the methods of each embodiment of the present application. The aforementioned storage medium includes: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program code.

[0164] Those skilled in the art will clearly understand that for the sake of convenience and brevity, the division of the above-mentioned functional modules is only used as an example for illustration. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the device described above can refer to the corresponding process in the aforementioned method embodiment and will not be repeated here.

[0165] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present application that follow the general principles of the present application and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, and the true scope and spirit of the present application are indicated by the following claims.

[0166] It should be understood that the present application is not limited to the exact structure described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

Claims

1. A method for intelligently entering a form, characterized in that: include: In response to an entry start operation on a first field in a form, obtaining content information of a second field entered in the form; Inputting the content information of the second field into the association model, and if it is determined through the association model that the number of records of the historical content information of the first field is greater than or equal to a field record count threshold, and the diversity parameter of the first field is greater than or equal to a field diversity threshold, determining first recommended content for the first field based on the content information of the second field and the association information between the second field and the first field; In response to an input operation on a first field in a form, matching the input content with the first recommended content of the first field according to the input content of the first field, and determining second recommended content that matches the input content; outputting second recommended content matching the input content; In response to a selection operation on any of the second recommended contents, automatically entering the selected second recommended content into the entry box of the first field; Before obtaining the content information of the second field entered in the form in response to the entry start operation of the first field in the form, the method further includes: In response to an import operation on a form file of the form, obtaining the imported form file, wherein form content information contained in the form file is inconsistent with the structure of the specified template of the form; Performing identification processing on the form file to determine the text content of the form file and the form type of the form; Structuring the text content of the form file according to a specified template corresponding to the form type of the form, and determining content information of at least one field of the form file; The content information of the at least one field is automatically entered into the corresponding entry area.

2. The method according to claim 1, characterized in that Also includes: Get the popular content information of each field contained in each form type; The association model is trained based on the popular content information of each field contained in each form type, and the field record count threshold and the field diversity threshold are determined. The trained association model includes candidate content information of each field contained in each form type, and at least one piece of association information, wherein each piece of association information includes a mapping rule between at least one input field and an output field, and is used to determine the candidate content information of the output field that meets the mapping rule based on the content information of the at least one input field.

3. The method according to claim 2, characterized in that The method of obtaining the popular content information of each field contained in each form type includes: Get the imported history form file; Performing identification processing on each of the historical form files to determine the form type corresponding to the historical form file and historical content information of each field in the historical form file; Aggregate historical content information of the same field based on the field type to determine the historical content information that each field has appeared in and the frequency of occurrence of each historical content information; According to the occurrence frequency information of each historical content information of each field, the popular content information of each field is determined.

4. The method according to claim 3, characterized in that Also includes: In response to a submission operation on the form, determining a recommendation feedback result for each recommended content in each field of the form, the recommendation feedback result being either adopted or not adopted; Determine the number of times each candidate content information in each field is adopted based on the recommendation feedback result of each recommended content; The priority recommendation order of the candidate content information of each field is updated according to the number of times each piece of candidate content information of each field is adopted.

5. An electronic device, characterized in that: include: a processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of claims 1 to 4.

6. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 4 when executed by a processor.

7. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the computer program is used to implement the method according to any one of claims 1 to 4.

Images