Processing method and processing system for automatically complementing form

Abstract

The embodiment of the invention relates to the technical field of artificial intelligence, and discloses a processing method and a processing system for automatically complementing a form, and the method comprises the following steps: matching similar form data from a form data knowledge base, wherein the similarity between the similar form data and a current form theme and a current content recognition field in currently filled form information reaches a specified threshold value; according to the current filled form information and the similar form data, form completion context information and form completion prompt information are generated respectively; and inputting the form completion context information and the form completion prompt information into a pre-trained Chain Of Though model based on a Dspy framework, so as to obtain a completion scheme of the current residual form. By implementing the embodiment of the invention, the accuracy and efficiency of form processing can be improved, and the user experience and satisfaction can be improved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a processing method and a processing system for automatic form completion. Background Art

[0002] At present, the development of user interface (UI) is changing with each passing day. With the advancement of technology and the improvement of user experience requirements, user interface (UI) design pays more and more attention to interactivity, intelligence and personalization. However, in practice, it is found that most of the existing user interfaces (UI) not only have the problem that the collection and analysis of user behavior data is not accurate and comprehensive enough, and it is difficult to deeply explore user needs and behavior patterns, but also the form processing process is cumbersome and error-prone, which reduces the user experience and efficiency. Summary of the invention

[0003] The embodiment of the present invention discloses a processing method and a processing system for automatic form completion, which can improve the accuracy and efficiency of form processing and enhance user experience and satisfaction.

[0004] A first aspect of an embodiment of the present invention discloses a method for processing form automatic completion, the method comprising: Matching similar form data from the form data knowledge base, the similarity of which reaches a specified threshold with the current form subject and the current content identification field in the currently filled-in form information; Generate form completion context information and form completion prompt information respectively according to the currently filled-in form information and the similar form data; The form completion context information and the form completion prompt information are input into a pre-trained Chain Of Thought model based on the Dspy framework to obtain a completion solution for the current remaining form.

[0005] As another optional implementation, in the first aspect of the embodiment of the present invention, before matching similar form data whose similarity with the current form subject and the current content identification field in the currently filled-in form information reaches a specified threshold from the form data knowledge base, the method further includes: Use SDK technology to set the basic information of the form and configure the form change monitoring function; wherein the basic information of the form includes the form subject and content identification fields; When the form change monitoring function monitors that the basic form information has changed, the user input behavior data is obtained, and the form filling event formed by the basic form information and the user input behavior data is reported to the form database; wherein the user input behavior data includes the name of the modified field and the filled content; The form filling events in the form database are processed by subject grouping of form data and merging of form data fields to determine final form data, and the final form data is stored in the form knowledge base.

[0006] As another optional implementation, in the first aspect of the embodiment of the present invention, the subject grouping of form data and the merging of form data fields of the form filling events in the form database to determine the final form data, and storing the final form data in the form knowledge base, includes: At regular intervals, incrementally extracting form filling events from the form database; After grouping the form data according to the subject, the last modified field in the form data field is retained to determine the final identification field of the form data; Final form data is determined according to the form data subject and the final identification field of the form data, and the final form data is stored in the form knowledge base.

[0007] As another optional implementation, in the first aspect of the embodiment of the present invention, matching similar form data whose similarity with the current form subject and the current content identification field in the currently filled-in form information reaches a specified threshold from the form data knowledge base includes: Combine the current form subject and the current content identification field to obtain content to be matched; Calculating the similarity between the processed form data in the form data knowledge base and the content to be matched using a Text Embedding model; The processed form data whose similarity reaches the specified threshold is used as the similar form data.

[0008] A second aspect of an embodiment of the present invention discloses a processing system, the processing system comprising: A matching unit, used to match similar form data from a form data knowledge base whose similarity with the current form subject and the current content identification field in the currently filled-in form information reaches a specified threshold; A generating unit, configured to generate form completion context information and form completion prompt information respectively according to the currently filled-in form information and the similar form data; The input unit is used to input the form completion context information and the form completion prompt information into a pre-trained Chain Of Thought model based on the Dspy framework to obtain a completion solution for the current remaining form.

[0009] As another optional implementation, in the second aspect of the embodiment of the present invention, the processing system includes: A configuration unit, used for setting the basic form information using SDK technology and configuring the form change monitoring function before the matching unit matches similar form data whose similarity with the current form subject and the current content identification field in the currently filled form information reaches a specified threshold from the form data knowledge base; wherein the basic form information includes the form subject and the content identification field; An acquisition and input unit, used for acquiring user input behavior data when the form change monitoring function monitors that the form basic information has changed, and reporting a form filling event formed by the form basic information and the user input behavior data to a form database; wherein the user input behavior data includes the name of the modified field and the filled content; The processing and storage unit is used to perform form data subject grouping and form data field merging processing on the form filling events in the form database to determine the final form data, and store the final form data in the form knowledge base.

[0010] As another optional implementation, in the second aspect of the embodiment of the present invention, the processing and storage unit includes: An extraction subunit, used for incrementally extracting form filling events from the form database at intervals of a certain period of time; A retaining subunit is used to retain the last modified field in the form data field after grouping and processing according to the subject of the form data, so as to determine the final identification field of the form data; The determination and storage subunit is used to determine the final form data according to the form data subject and the final identification field of the form data, and store the final form data in the form knowledge base.

[0011] As another optional implementation, in the second aspect of the embodiment of the present invention, the matching unit includes: A combining subunit, used for combining the current form subject and the current content identification field to obtain content to be matched; A calculation subunit, used to calculate the similarity between the processed form data in the form data knowledge base and the content to be matched by using a Text Embedding model; The determination subunit is used to take the processed form data whose similarity reaches the specified threshold as the similar form data.

[0012] A third aspect of an embodiment of the present invention discloses a processing system, the processing system comprising: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute a form automatic completion processing method disclosed in the first aspect of an embodiment of the present invention.

[0013] A fourth aspect of an embodiment of the present invention discloses a computer-readable storage medium storing a computer program, wherein the computer program enables a computer to execute a form automatic completion processing method disclosed in the first aspect of an embodiment of the present invention.

[0014] A fifth aspect of an embodiment of the present invention discloses a computer program product. When the computer program product is run on a computer, the computer is enabled to execute part or all of the steps of any one of the processing methods for automatic form completion in the first aspect.

[0015] A sixth aspect of an embodiment of the present invention discloses an application publishing platform, which is used to publish a computer program product. When the computer program product runs on a computer, the computer executes part or all of the steps of any one of the form automatic completion processing methods of the first aspect.

[0016] Compared with the prior art, the embodiments of the present invention have the following beneficial effects: In an embodiment of the present invention, similar form data whose similarity with the current form subject and the current content identification field in the currently filled-in form information reaches a specified threshold is matched from the form data knowledge base; form completion context information and form completion prompt information are generated respectively according to the currently filled-in form information and the similar form data; the form completion context information and the form completion prompt information are input into a pre-trained Chain Of Thought model based on the Dspy framework to obtain a completion solution for the current remaining form. It can be seen that the embodiment of the present invention can improve the accuracy and efficiency of form processing and enhance user experience and satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 It is a flowchart of a method for automatically completing a form disclosed in an embodiment of the present invention; Figure 2 It is a flowchart of another form automatic completion processing method disclosed in an embodiment of the present invention; Figure 3 It is a structural schematic diagram of a processing system disclosed in an embodiment of the present invention; Figure 4 is a structural schematic diagram of another processing system disclosed in an embodiment of the present invention; Figure 5 It is a structural schematic diagram of another processing system disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0019] 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 only 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.

[0020] It should be noted that the terms "first", "second", "third", "fourth", etc. in the specification and claims of the present invention are used to distinguish different objects rather than to describe a specific order. The terms "including" and "having" in the embodiments of the present invention and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device including a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0021] The embodiment of the present invention discloses a processing method and a processing system for automatic form completion, which can improve the accuracy and efficiency of form processing and enhance user experience and satisfaction.

[0022] The following is a detailed description with reference to the accompanying drawings. Embodiment 1

[0023] See also Figure 1 , Figure 1 1 is a flow chart of a method for automatically completing a form disclosed in an embodiment of the present invention. Figure 1 As shown, the processing method for automatically completing the form may include the following steps.

[0024] 101. The processing system matches similar form data from a form data knowledge base whose similarity with the current form subject and the current content identification field in the currently filled-in form information reaches a specified threshold.

[0025] As an optional implementation, in an embodiment of the present invention, first, the system utilizes a form data knowledge base, which is a data warehouse that stores a large amount of historical form data. By mining and utilizing these historical data, the system can provide experience-based recommendations and completions, which are not available in traditional form filling systems. Secondly, the system can identify fields based on the subject and content of the current form, and search for form data in the knowledge base whose similarity reaches a specified threshold. This matching not only relies on simple keyword or field matching, but may also involve more complex semantic analysis or machine learning algorithms, thereby improving the accuracy and practicality of the matching. Finally, by matching similar form data, the system can provide each user with personalized recommendations based on the content of the form they are currently filling out, thereby improving user experience and satisfaction.

[0026] As an optional implementation, in an embodiment of the present invention, by matching the historical form data in the form data knowledge base, the system can quickly locate historical events similar to the current form filling situation. At the same time, the system can use historical experience to guide the current form filling, which can effectively improve the accuracy and efficiency of form completion.

[0027] As an optional implementation, in an embodiment of the present invention, the present application uses the form filling mode in the historical data to provide a reference for the current form filling. At the same time, by matching similar form data, basic data and context information can be provided for subsequent form completion.

[0028] As an optional implementation, in an embodiment of the present invention, the system can adopt advanced natural language processing and machine learning algorithms, which are not only based on simple keyword matching, but also consider complex factors such as contextual semantics, word order, and syntactic structure, so as to more accurately match historical form data that is highly similar to the current form information from the form data knowledge base.

[0029] As an optional implementation, in an embodiment of the present invention, in addition to traditional evaluation methods such as cosine similarity and Jaccard similarity, the system can also introduce deep learning-based embedding representation methods (such as BERT, GPT, etc.) to calculate the semantic similarity between texts to ensure the accuracy and diversity of the matching results.

[0030] As an optional implementation, in an embodiment of the present invention, the system can dynamically adjust the similarity threshold according to the complexity of the current form information and the richness of the historical form data to balance the matching accuracy and recall rate and improve the overall performance of the system.

[0031] 102. The processing system generates form completion context information and form completion prompt information respectively according to the currently filled-in form information and similar form data.

[0032] As an optional implementation, in an embodiment of the present invention, first, the system can generate context information that helps users understand the background and requirements of form filling based on the currently filled form information and similar form data. The generation of such information requires a deep understanding of the structure and content of the form, as well as the relationship between them. Secondly, the system can intelligently push completion prompt information related to the current form filling progress. These prompt information may include filling options, format requirements, filling suggestions, etc. This intelligent push can significantly improve the user's filling efficiency and accuracy. Finally, by providing context information and prompt information, the system makes the form filling process more intuitive and easy to understand, reduces the user's cognitive burden, and improves the user's experience.

[0033] As an optional implementation, in an embodiment of the present invention, the generated form completion context information provides detailed background information for subsequent completion solutions. At the same time, the form completion prompt information directly gives suggestions for the difficulty or missing parts of the current form, helping users complete the form faster.

[0034] As an optional implementation, in the embodiment of the present invention, the present application constructs a complete form filling environment, including historical experience and current status, and also provides specific completion suggestions to reduce the difficulty and error rate of users filling out forms.

[0035] As an optional implementation, in an embodiment of the present invention, the system can generate personalized form completion context information by analyzing the similarities and differences between the currently filled-in form information and similar form data, so as to help users better understand the requirements and background of form filling.

[0036] As an optional implementation, in an embodiment of the present invention, based on the predictive ability of the machine learning model, the system can intelligently push completion prompt information that is highly relevant to the current form filling progress, such as possible filling options, filling format requirements, etc., to improve the user's filling efficiency and accuracy.

[0037] As an optional implementation, in an embodiment of the present invention, the system can also support interactive operations between users and completion prompt information, such as quickly completing form filling by clicking, dragging, etc., to further enhance the user experience.

[0038] 103. The processing system inputs the form completion context information and the form completion prompt information into a pre-trained Chain Of Thought model based on the Dspy framework to obtain a completion solution for the current remaining form.

[0039] As an optional implementation, in an embodiment of the present invention, first, the Chain Of Thought model is an advanced machine learning model that can generate complex outputs through multi-step reasoning. In this scenario, it is used to generate a form completion solution, which is a breakthrough in traditional form completion methods. Secondly, the Chain Of Thought model can perform multi-step reasoning and analysis based on the input context information and prompt information to generate a complete form completion solution. This capability enables the system to handle more complex form filling tasks, improving the practicality and flexibility of the system. Finally, by utilizing the powerful reasoning ability of the Chain Of Thought model, the system can efficiently generate accurate form completion solutions, thereby reducing the user's filling time and error rate and improving overall work efficiency.

[0040] As an optional implementation, in an embodiment of the present invention, by utilizing the Chain Of Thought model based on the Dspy framework, the system can generate a more accurate and reasonable form completion solution based on context information and prompt information, while also improving the intelligence level of form completion, enabling the system to handle more complex and diverse form filling scenarios.

[0041] As an optional implementation, in an embodiment of the present invention, this application effectively realizes the intelligent and automated form completion through the powerful capabilities of the machine learning model, while also providing an efficient and accurate form completion solution to enhance the user experience and satisfaction.

[0042] As an optional implementation, in an embodiment of the present invention, the system achieves efficient, accurate and intelligent form completion by matching similar form data from a form data knowledge base, generating form completion context information and prompt information, and using a Chain Of Thought model based on a Dspy framework to generate a completion solution. This can reduce the difficulty and error rate of users filling out forms and improve user experience and satisfaction.

[0043] As an optional implementation, in an embodiment of the present invention, the Chain Of Thought model adopted by the system is a customized model based on the Dspy framework. By introducing advanced technologies such as attention mechanism and multi-head self-attention, it can have a deeper understanding of the logical relationships and dependencies in the form filling process, thereby generating a more reasonable and complete form completion solution.

[0044] As an optional implementation, in an embodiment of the present invention, the Chain Of Thought model has multi-step reasoning capabilities, and can gradually deduce the filling content and order of the current remaining forms based on the form completion context information and prompt information, thereby ensuring the accuracy and completeness of the completion plan.

[0045] As an optional implementation, in an embodiment of the present invention, the system supports real-time receipt of user feedback and adjustments during the process of generating completion solutions, continuously optimizing and improving the completion solutions through the online learning mechanism of the machine learning model, and improving the intelligence level of the system and user satisfaction.

[0046] As an optional implementation, in an embodiment of the present invention, before using the Chain Of Thought model of DSPY, the developer needs to prepare several cases similar to actual application scenarios. These cases should contain complete input, intermediate reasoning steps and final output. These cases will be used to train the model so that it learns how to extract key information from the input, build a logical chain, generate a reasonable reasoning chain, and finally draw accurate conclusions.

[0047] As an optional implementation, in an embodiment of the present invention, the present application may prepare several examples of automatic completion before executing step 103, each example including four parts: context, question, hint, and answer. Context is the text to be completed, which contains one or more spaces or placeholders, indicating the part that needs to be automatically completed. Question clearly indicates the content or location that needs to be completed. Hint provides reference information related to the content to be completed, which may be a similar text or context. Answer is the complete text or answer after completion.

[0048] As an optional implementation, in an embodiment of the present invention, after the Chain Of Thought model is trained, it can be used in a production environment to improve the accuracy and efficiency of form processing. First, the system can construct a context (Context) containing this information based on the partial form information filled in by the user. This context will be used as the input of the ChainOfThought model to generate subsequent logical reasoning and answers. Subsequently, the system will find processed forms similar to the current form from the form knowledge base based on the subject and content of the current form. Using the information of these similar forms, a hint (Hint) is constructed, which will contain some additional information that may help the model generate accurate answers. The purpose of this step is to provide the model with a broader context to help it better understand and process the current form. Finally, the system can use the constructed Context and Hint as input to call the trained Chain Of Thought model. The model will generate an output (Answer) containing logical reasoning steps and the final answer based on the input information. This answer is displayed to the user, and if the user chooses to accept it, the answer is automatically filled in the corresponding form item.

[0049] As an optional implementation, in an embodiment of the present invention, this application can significantly improve the accuracy and efficiency of form processing by combining the DSPY framework with the ChainOf Thought model and utilizing the powerful capabilities of the large model.

[0050] exist Figure 1 In the processing method of the form automatic completion, the processing system is used as an example to describe it. It should be noted that Figure 1 The execution subject of the form automatic completion processing method may also be an independent device associated with the processing system, which is not limited in the embodiment of the present invention.

[0051] It can be seen that implementation Figure 1 The described method for automatically completing a form can improve the accuracy and efficiency of form processing and enhance user experience and satisfaction.

[0052] In addition, implementation Figure 1 The described method for automatically completing a form can make the form processing process more intelligent and automated, and reduce manual intervention and error rate. Embodiment 2

[0053] See also Figure 2 , Figure 2 FIG. 1 is a flow chart of another method for automatically completing a form disclosed in an embodiment of the present invention. Figure 2 , the processing method of the form automatic completion may include the following steps: 201. The processing system uses SDK technology to set basic form information and configures a form change monitoring function; wherein the basic form information includes a form subject and a content identification field.

[0054] As an optional implementation, in the embodiment of the present invention, the SDK provided by the system not only supports basic form information settings, but also allows developers to highly customize form fields, validation rules, etc. according to actual needs. In addition, the SDK also integrates a powerful form change monitoring mechanism that can capture user input behavior in real time, providing a basis for subsequent data processing and event reporting. Moreover, the SDK has been carefully designed to be seamlessly integrated into a variety of front-end frameworks and platforms, including Web, mobile terminals, etc., to achieve a consistent experience across platforms.

[0055] As an optional implementation, in an embodiment of the present invention, this application uses SDK technology, and the processing system can flexibly set the basic information of the form and expand the function according to the needs. This technology not only improves the customizability of the system, but also ensures that the system can easily adapt to the form requirements in different scenarios. At the same time, by configuring the form change monitoring function, the system can capture the changes in the basic information of the form in real time. This real-time performance is crucial to improving the accuracy and timeliness of data processing, and provides a reliable foundation for subsequent form data reporting and processing.

[0056] As an optional implementation, in the embodiment of the present invention, the present application utilizes SDK (Software Development Kit) technology, which not only simplifies the process of setting the basic information of the form, but also realizes real-time monitoring of form changes. This integrated method improves the flexibility and response speed of form processing, allowing the system to capture the user's input behavior in real time.

[0057] 202. When the form change monitoring function detects that the basic information of the form has changed, the processing system obtains the user input behavior data and reports the form filling event formed by the basic information of the form and the user input behavior data to the form database; wherein the user input behavior data includes the name of the modified field and the filled content.

[0058] As an optional implementation, in the embodiment of the present invention, the system can capture and report every input behavior of the user, including modifying the field name, filling in the content, etc. This fine-grained data collection provides a rich source of information for subsequent data analysis and form completion. At the same time, the system adopts a real-time data reporting mechanism to ensure that the user input behavior data can be captured and processed by the system in real time, improving the response speed and accuracy of the system.

[0059] As an optional implementation, in the embodiment of the present invention, the system can obtain and report user input behavior data, including the name of the modified field and the filled content. These data provide the system with rich user behavior information, which helps to deeply understand user needs and form filling habits. At the same time, through the analysis of user behavior data, the system can find the pain points and bottlenecks in the form filling process, so as to carry out targeted optimization. This data-driven approach not only improves the usability of the form, but also improves the user experience.

[0060] As an optional implementation, in the embodiment of the present invention, the system captures user input behavior data (such as the name of the modified field and the filled content) through the monitoring function, and combines it with the basic information of the form to form a form filling event and report it to the database. This approach realizes the fine-grained recording and analysis of user operation behavior, providing a rich data source for subsequent data processing and intelligent completion.

[0061] 203. At regular intervals, the processing system incrementally extracts form filling events from the form database.

[0062] As an optional implementation, in the embodiment of the present invention, the system uses incremental data extraction technology to extract only the form filling events added since the last processing, avoiding repeated processing and improving system efficiency. This incremental processing method is particularly suitable for large-scale data sets and ensures the performance and stability of the system.

[0063] 204. After grouping the form data according to the subject, the processing system retains the last modified field in the form data field to determine the final identification field of the form data.

[0064] As an optional implementation, in an embodiment of the present invention, the system can automatically perform grouping processing and determine the final identification field based on the subject and field modification of the form data. This intelligent processing method not only improves the accuracy and efficiency of data processing, but also reduces the cost of manual intervention.

[0065] 205. The processing system determines the final form data according to the form data subject and the final identification field of the form data, and stores the final form data in the form knowledge base.

[0066] As an optional implementation, in the embodiment of the present invention, by determining the final form data and storing it in the form knowledge base, the system realizes dynamic updating and enrichment of form knowledge. This dynamic updating mechanism of the knowledge base provides more comprehensive and accurate data support for subsequent form completion and recommendation.

[0067] As an optional implementation, in the embodiment of the present invention, the present application uses an incremental extraction method to extract data from the form database, reducing the complexity of data processing. At the same time, by grouping and retaining the last modified field, the final identification field of the form data is determined, which helps to reduce redundant information and improve the accuracy and efficiency of data processing.

[0068] 206. The processing system combines the current form subject and the current content identification field to obtain content to be matched.

[0069] 207. The processing system uses the Text Embedding model to calculate the similarity between the processed form data in the form data knowledge base and the content to be matched.

[0070] 208. The processing system uses the processed form data whose similarity reaches a specified threshold as similar form data.

[0071] As an optional implementation, in an embodiment of the present invention, the system uses the Text Embedding model to calculate the similarity between the content to be matched and the processed form data, thereby achieving accurate matching and personalized recommendations. This accurate matching mechanism not only improves the accuracy and efficiency of form completion, but also enhances the user experience and satisfaction. At the same time, the application of the Text Embedding model enables the system to calculate the similarity in a short time and achieve rapid response. The application of this efficient algorithm greatly improves the performance of the system and the user experience.

[0072] As an optional implementation, in the embodiment of the present invention, the system uses the Text Embedding model to calculate the similarity between the content to be matched and the processed form data, which is an advanced natural language processing technology that can capture the semantic similarity between texts. This technology enables the system to intelligently identify historical events similar to the current form, providing a valuable reference for form completion.

[0073] 209. The processing system generates form completion context information and form completion prompt information respectively according to the currently filled-in form information and similar form data.

[0074] As an optional implementation, in an embodiment of the present invention, the system generates personalized form completion context information and prompt information based on the currently filled-in form information and similar form data. This personalized completion suggestion not only improves the efficiency and accuracy of form filling, but also reduces the user's cognitive burden and filling difficulty. At the same time, by generating form completion context information and prompt information, the system can provide users with intelligent assistance and guidance to help users complete form filling faster. This intelligent assistance mechanism greatly improves the user experience and satisfaction.

[0075] 210. The processing system inputs the form completion context information and the form completion prompt information into the pre-trained Chain Of Thought model based on the Dspy framework to obtain the completion solution for the current remaining form, and ends this process.

[0076] As an optional implementation, in an embodiment of the present invention, the system can intelligently generate a completion solution for the current remaining form based on the form completion context information and prompt information through a pre-trained Chain Of Thought model based on the Dspy framework. This intelligent completion solution not only improves the accuracy and efficiency of form filling, but also reduces the difficulty and cognitive burden of users. At the same time, from the setting of basic form information to the generation of the final completion solution, the system provides a complete end-to-end solution. This solution not only improves the overall performance and stability of the system, but also ensures the smoothness and continuity of the form filling process. At the same time, by continuously optimizing and upgrading the Chain Of Thought model, the system can continuously improve the intelligence and accuracy of form completion.

[0077] As an optional implementation, in the embodiment of the present invention, the system introduces the ChainOf Thought model based on the Dspy framework, which can generate the completion solution for the current remaining form based on the form completion context information and the form completion prompt information. The ChainOf Thought model, with its powerful logical reasoning and context understanding capabilities, has demonstrated outstanding performance in the field of intelligent completion and improved the efficiency and accuracy of form filling.

[0078] As an optional implementation method, in the embodiment of the present invention, the entire process of this application from form setting, monitoring, data reporting, incremental processing to intelligent completion is highly automated and intelligent. This innovative process design not only improves the efficiency of form processing, but also significantly improves the user experience, making form filling more convenient and intelligent.

[0079] It can be seen that implementation Figure 2 Another form auto-completion processing method described herein can improve the accuracy and efficiency of form processing and enhance user experience and satisfaction.

[0080] In addition, implementation Figure 2 Another form auto-completion processing method described can make form filling more convenient and intelligent. Embodiment 3

[0081] See also Figure 3 , Figure 3 Schematic diagram of a processing system disclosed in an embodiment of the present invention. Figure 3 , the processing system 300 may include a matching unit 301, a generating unit 302 and an input unit 303, wherein: The matching unit 301 is used to match similar form data from the form data knowledge base, the similarity of which reaches a specified threshold with the current form subject and the current content identification field in the currently filled-in form information.

[0082] The generating unit 302 is used to generate form completion context information and form completion prompt information according to the currently filled-in form information and similar form data.

[0083] The input unit 303 is used to input the form completion context information and the form completion prompt information into a pre-trained Chain Of Thought model based on the Dspy framework to obtain a completion solution for the current remaining form.

[0084] As an optional implementation, in an embodiment of the present invention, first, the system utilizes a form data knowledge base, which is a data warehouse that stores a large amount of historical form data. By mining and utilizing these historical data, the system can provide experience-based recommendations and completions, which are not available in traditional form filling systems. Secondly, the system can identify fields based on the subject and content of the current form, and search for form data in the knowledge base whose similarity reaches a specified threshold. This matching not only relies on simple keyword or field matching, but may also involve more complex semantic analysis or machine learning algorithms, thereby improving the accuracy and practicality of the matching. Finally, by matching similar form data, the system can provide each user with personalized recommendations based on the content of the form they are currently filling out, thereby improving user experience and satisfaction.

[0085] As an optional implementation, in an embodiment of the present invention, the matching unit 301 matches the historical form data in the form data knowledge base, and the system can quickly locate historical events similar to the current form filling situation. At the same time, the system can use historical experience to guide the current form filling, which can effectively improve the accuracy and efficiency of form completion.

[0086] As an optional implementation, in an embodiment of the present invention, the present application uses the form filling mode in the historical data to provide a reference for the current form filling. At the same time, the matching unit 301 can provide basic data and context information for subsequent form completion by matching similar form data.

[0087] As an optional implementation, in an embodiment of the present invention, the system can adopt advanced natural language processing and machine learning algorithms, which are not only based on simple keyword matching, but also consider complex factors such as contextual semantics, word order, and syntactic structure, so as to more accurately match historical form data that is highly similar to the current form information from the form data knowledge base.

[0088] As an optional implementation, in an embodiment of the present invention, in addition to traditional evaluation methods such as cosine similarity and Jaccard similarity, the system can also introduce deep learning-based embedding representation methods (such as BERT, GPT, etc.) to calculate the semantic similarity between texts to ensure the accuracy and diversity of the matching results.

[0089] As an optional implementation, in an embodiment of the present invention, the system can dynamically adjust the similarity threshold according to the complexity of the current form information and the richness of the historical form data to balance the matching accuracy and recall rate and improve the overall performance of the system.

[0090] As an optional implementation, in an embodiment of the present invention, first, the system can generate context information that helps users understand the background and requirements of form filling based on the currently filled form information and similar form data. The generation of such information requires a deep understanding of the structure and content of the form, as well as the relationship between them. Secondly, the system can intelligently push completion prompt information related to the current form filling progress. These prompt information may include filling options, format requirements, filling suggestions, etc. This intelligent push can significantly improve the user's filling efficiency and accuracy. Finally, by providing context information and prompt information, the system makes the form filling process more intuitive and easy to understand, reduces the user's cognitive burden, and improves the user's experience.

[0091] As an optional implementation, in the embodiment of the present invention, the form completion context information generated by the generation unit 302 provides detailed background information for subsequent completion solutions. At the same time, the form completion prompt information directly gives suggestions for the difficulty or missing parts of the current form, helping the user to complete the form faster.

[0092] As an optional implementation, in the embodiment of the present invention, the generating unit 302 constructs a complete form filling environment, including historical experience and current status, and also provides specific completion suggestions to reduce the difficulty and error rate of users filling out forms.

[0093] As an optional implementation, in an embodiment of the present invention, the generation unit 302 can generate personalized form completion context information by analyzing the similarities and differences between the currently filled-in form information and similar form data, so as to help users better understand the requirements and background of form filling.

[0094] As an optional implementation, in an embodiment of the present invention, based on the predictive ability of the machine learning model, the generation unit 302 can intelligently push completion prompt information that is highly relevant to the current form filling progress, such as possible filling options, filling format requirements, etc., to improve the user's filling efficiency and accuracy.

[0095] As an optional implementation, in an embodiment of the present invention, the system can also support interactive operations between users and completion prompt information, such as quickly completing form filling by clicking, dragging, etc., to further enhance the user experience.

[0096] As an optional implementation, in an embodiment of the present invention, first, the Chain Of Thought model is an advanced machine learning model that can generate complex outputs through multi-step reasoning. In this scenario, it is used to generate a form completion solution, which is a breakthrough in traditional form completion methods. Secondly, the Chain Of Thought model can perform multi-step reasoning and analysis based on the input context information and prompt information to generate a complete form completion solution. This capability enables the system to handle more complex form filling tasks, improving the practicality and flexibility of the system. Finally, by utilizing the powerful reasoning ability of the Chain Of Thought model, the system can efficiently generate accurate form completion solutions, thereby reducing the user's filling time and error rate and improving overall work efficiency.

[0097] As an optional implementation, in an embodiment of the present invention, by utilizing the Chain Of Thought model based on the Dspy framework, the system can generate a more accurate and reasonable form completion solution based on context information and prompt information, while also improving the intelligence level of form completion, enabling the system to handle more complex and diverse form filling scenarios.

[0098] As an optional implementation, in an embodiment of the present invention, this application effectively realizes the intelligent and automated form completion through the powerful capabilities of the machine learning model, while also providing an efficient and accurate form completion solution to enhance the user experience and satisfaction.

[0099] As an optional implementation, in an embodiment of the present invention, the system achieves efficient, accurate and intelligent form completion by matching similar form data from a form data knowledge base, generating form completion context information and prompt information, and using a Chain Of Thought model based on a Dspy framework to generate a completion solution. This can reduce the difficulty and error rate of users filling out forms and improve user experience and satisfaction.

[0100] As an optional implementation, in an embodiment of the present invention, the Chain Of Thought model adopted by the system is a customized model based on the Dspy framework. By introducing advanced technologies such as attention mechanism and multi-head self-attention, it can have a deeper understanding of the logical relationships and dependencies in the form filling process, thereby generating a more reasonable and complete form completion solution.

[0101] As an optional implementation, in an embodiment of the present invention, the Chain Of Thought model has multi-step reasoning capabilities, and can gradually deduce the filling content and order of the current remaining forms based on the form completion context information and prompt information, thereby ensuring the accuracy and completeness of the completion plan.

[0102] As an optional implementation, in an embodiment of the present invention, the system supports real-time receipt of user feedback and adjustments during the process of generating completion solutions, continuously optimizing and improving the completion solutions through the online learning mechanism of the machine learning model, and improving the intelligence level of the system and user satisfaction.

[0103] As an optional implementation, in an embodiment of the present invention, before using the Chain Of Thought model of DSPY, the developer needs to prepare several cases similar to actual application scenarios. These cases should contain complete input, intermediate reasoning steps and final output. These cases will be used to train the model so that it learns how to extract key information from the input, build a logical chain, generate a reasonable reasoning chain, and finally draw accurate conclusions.

[0104] As an optional implementation, in an embodiment of the present invention, the input unit 303 may prepare several examples of automatic completion before execution, each example including four parts: context, question, hint and answer. Context is the text to be completed, which contains one or more spaces or placeholders, indicating the part that needs to be automatically completed. Question clearly indicates the content or location that needs to be completed. Hint provides reference information related to the content to be completed, which may be a similar text or context. Answer is the complete text or answer after completion.

[0105] As an optional implementation, in an embodiment of the present invention, after the Chain Of Thought model is trained, it can be used in a production environment to improve the accuracy and efficiency of form processing. First, the system can construct a context (Context) containing this information based on the partial form information filled in by the user. This context will be used as the input of the ChainOfThought model to generate subsequent logical reasoning and answers. Subsequently, the matching unit 301 will find processed forms similar to the current form from the form knowledge base based on the subject and content of the current form. Using the information of these similar forms, the generation unit 302 constructs a hint (Hint), which will contain some additional information that may help the model generate accurate answers. The purpose of this step is to provide the model with a broader context to help it better understand and process the current form. Finally, the input unit 303 can use the constructed Context and Hint as input to call the trained Chain Of Thought model. The model will generate an output (Answer) containing logical reasoning steps and the final answer based on the input information. This answer is displayed to the user, and if the user chooses to accept, the answer is automatically filled in the corresponding form item.

[0106] As an optional implementation, in an embodiment of the present invention, this application can significantly improve the accuracy and efficiency of form processing by combining the DSPY framework with the ChainOf Thought model and utilizing the powerful capabilities of the large model.

[0107] It can be seen that implementation Figure 3 The processing system described can improve the accuracy and efficiency of form processing and enhance user experience and satisfaction.

[0108] In addition, implementation Figure 3 The processing system described can make the form processing process more intelligent and automated, reducing manual intervention and error rates. Embodiment 4

[0109] See also Figure 4 , Figure 4 is a schematic diagram of the structure of another processing system disclosed in an embodiment of the present invention. Figure 4 The processing system is composed of Figure 3 The processing system is optimized. Figure 3 Compared with the processing system, Figure 4 The processing system includes: The configuration unit 304 is used to set the basic form information using SDK technology and configure the form change listening function before the matching unit 301 matches similar form data whose similarity with the current form subject and the current content identification field in the currently filled-in form information reaches a specified threshold from the form data knowledge base; wherein the basic form information includes the form subject and the content identification field.

[0110] As an optional implementation, in the embodiment of the present invention, the SDK provided by the configuration unit 304 not only supports basic form information settings, but also allows developers to highly customize form fields, validation rules, etc. according to actual needs. In addition, the SDK also integrates a powerful form change monitoring mechanism, which can capture user input behavior in real time, providing a basis for subsequent data processing and event reporting. Moreover, the SDK has been carefully designed to be seamlessly integrated into a variety of front-end frameworks and platforms, including Web, mobile terminals, etc., to achieve a consistent experience across platforms.

[0111] As an optional implementation, in an embodiment of the present invention, the configuration unit 304 can flexibly set the basic information of the form through SDK technology and expand the function according to the needs. This technology not only improves the customizability of the system, but also ensures that the system can easily adapt to the form requirements in different scenarios. At the same time, by configuring the form change monitoring function, the configuration unit 304 can capture the changes in the basic information of the form in real time. This real-time performance is crucial to improving the accuracy and timeliness of data processing, and provides a reliable basis for subsequent form data reporting and processing.

[0112] As an optional implementation, in the embodiment of the present invention, the configuration unit 304 uses SDK (Software Development Kit) technology to not only simplify the process of setting the basic information of the form, but also realize real-time monitoring of form changes. This integrated method improves the flexibility and response speed of form processing, allowing the system to capture the user's input behavior in real time.

[0113] The acquisition and input unit 305 is used to acquire user input behavior data when the form change monitoring function monitors changes in the basic information of the form, and report the form filling event formed by the basic information of the form and the user input behavior data to the form database; wherein the user input behavior data includes the name of the modified field and the filled content.

[0114] As an optional implementation, in the embodiment of the present invention, the acquisition and input unit 305 can capture and report each user input behavior, including modifying field names, filling in content, etc. This fine-grained data collection provides a rich source of information for subsequent data analysis and form completion. At the same time, the acquisition and input unit 305 adopts a real-time data reporting mechanism to ensure that the user input behavior data can be captured and processed by the system in real time, thereby improving the response speed and accuracy of the system.

[0115] As an optional implementation, in an embodiment of the present invention, the acquisition and input unit 305 can acquire and report user input behavior data, including the name of the modified field and the filled content. These data provide the system with rich user behavior information, which helps to deeply understand user needs and form filling habits. At the same time, through the analysis of user behavior data, the system can find the pain points and bottlenecks in the form filling process, so as to perform targeted optimization. This data-driven method not only improves the usability of the form, but also improves the user experience.

[0116] As an optional implementation, in the embodiment of the present invention, the acquisition and input unit 305 captures the user input behavior data (such as the name of the modified field and the filled content) through the monitoring function, and combines it with the basic information of the form to form a form filling event and report it to the database. This approach realizes the fine-grained recording and analysis of user operation behavior, and provides a rich data source for subsequent data processing and intelligent completion.

[0117] The processing and storage unit 306 is used to perform form data subject grouping and form data field merging processing on the form filling events in the form database to determine the final form data, and store the final form data in the form knowledge base.

[0118] As an optional implementation, in the embodiment of the present invention, by determining the final form data and storing it in the form knowledge base, the processing and storage unit 306 realizes dynamic updating and enrichment of the form knowledge. This dynamic updating mechanism of the knowledge base provides more comprehensive and accurate data support for subsequent form completion and recommendation.

[0119] and Figure 3 Compared with the processing system, Figure 4 The processing and storage unit 306 includes: The extraction subunit 3061 is used to incrementally extract form filling events from the form database at regular intervals.

[0120] As an optional implementation, in the embodiment of the present invention, the extraction subunit 3061 adopts incremental data extraction technology to extract only the form filling events added since the last processing, thereby avoiding repeated processing and improving system efficiency. This incremental processing method is particularly suitable for large-scale data sets and ensures the performance and stability of the system.

[0121] The retaining subunit 3062 is used to retain the last modified field in the form data fields after grouping the form data according to the subject, so as to determine the final identification field of the form data.

[0122] As an optional implementation, in an embodiment of the present invention, the retention subunit 3062 can automatically perform grouping processing and determine the final identification field based on the subject and field modification of the form data. This intelligent processing method not only improves the accuracy and efficiency of data processing, but also reduces the cost of manual intervention.

[0123] The determination and storage subunit 3063 is used to determine the final form data according to the form data subject and the final identification field of the form data, and store the final form data in the form knowledge base.

[0124] As an optional implementation, in the embodiment of the present invention, the present application uses an incremental extraction method to extract data from the form database, reducing the complexity of data processing. At the same time, by grouping and retaining the last modified field, the final identification field of the form data is determined, which helps to reduce redundant information and improve the accuracy and efficiency of data processing.

[0125] and Figure 3 Compared with the processing system, Figure 4 The matching unit 301 comprises: The combining subunit 3011 is used to combine the current form subject and the current content identification field to obtain the content to be matched.

[0126] The calculation subunit 3012 is used to calculate the similarity between the processed form data in the form data knowledge base and the content to be matched by using the Text Embedding model.

[0127] The determination subunit 3013 is used to take the processed form data whose similarity reaches a specified threshold as similar form data.

[0128] As an optional implementation, in an embodiment of the present invention, the calculation subunit 3012 uses the TextEmbedding model to calculate the similarity between the content to be matched and the processed form data, thereby achieving accurate matching and personalized recommendation. This accurate matching mechanism not only improves the accuracy and efficiency of form completion, but also enhances the user experience and satisfaction. At the same time, the application of the Text Embedding model enables the system to calculate the similarity in a short time and achieve rapid response. The application of this efficient algorithm greatly improves the performance of the system and the user experience.

[0129] As an optional implementation, in the embodiment of the present invention, the calculation subunit 3012 uses the TextEmbedding model to calculate the similarity between the content to be matched and the processed form data, which is an advanced natural language processing technology that can capture the semantic similarity between texts. This technology enables the system to intelligently identify historical events similar to the current form, providing a valuable reference for form completion.

[0130] As an optional implementation, in an embodiment of the present invention, the system generates personalized form completion context information and prompt information based on the currently filled-in form information and similar form data. This personalized completion suggestion not only improves the efficiency and accuracy of form filling, but also reduces the user's cognitive burden and filling difficulty. At the same time, by generating form completion context information and prompt information, the system can provide users with intelligent assistance and guidance to help users complete form filling faster. This intelligent assistance mechanism greatly improves the user experience and satisfaction.

[0131] As an optional implementation, in an embodiment of the present invention, the system can intelligently generate a completion solution for the current remaining form based on the form completion context information and prompt information through a pre-trained Chain Of Thought model based on the Dspy framework. This intelligent completion solution not only improves the accuracy and efficiency of form filling, but also reduces the difficulty and cognitive burden of users. At the same time, from the setting of basic form information to the generation of the final completion solution, the system provides a complete end-to-end solution. This solution not only improves the overall performance and stability of the system, but also ensures the smoothness and continuity of the form filling process. At the same time, by continuously optimizing and upgrading the Chain Of Thought model, the system can continuously improve the intelligence and accuracy of form completion.

[0132] As an optional implementation, in the embodiment of the present invention, the system introduces the ChainOf Thought model based on the Dspy framework, which can generate the completion solution for the current remaining form based on the form completion context information and the form completion prompt information. The ChainOf Thought model, with its powerful logical reasoning and context understanding capabilities, has demonstrated outstanding performance in the field of intelligent completion and improved the efficiency and accuracy of form filling.

[0133] As an optional implementation method, in the embodiment of the present invention, the entire process of this application from form setting, monitoring, data reporting, incremental processing to intelligent completion is highly automated and intelligent. This innovative process design not only improves the efficiency of form processing, but also significantly improves the user experience, making form filling more convenient and intelligent.

[0134] It can be seen that implementation Figure 4 Another processing system described can improve the accuracy and efficiency of form processing and enhance user experience and satisfaction.

[0135] In addition, implementation Figure 4 Another processing system described can make form filling more convenient and intelligent. Embodiment 5

[0136] See also Figure 5 , Figure 5 FIG. 1 is a schematic diagram of another processing system disclosed in an embodiment of the present invention. Figure 5 , the processing system may include: A memory 501 storing executable program codes; a processor 502 coupled to the memory 501; The processor 502 calls the executable program code stored in the memory 501 to execute Figure 1~Figure 2 Any form autocomplete handler.

[0137] The embodiment of the present invention discloses a computer-readable storage medium storing a computer program, wherein the computer program enables a computer to execute Figure 1~Figure 2 Any form autocomplete handler.

[0138] The embodiments of the present invention further disclose a computer program product, wherein when the computer program product is run on a computer, the computer is enabled to execute part or all of the steps of the methods in the above method embodiments.

[0139] A person of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable programmable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0140] The above is a detailed introduction to a processing method and processing system for automatic form completion disclosed in an embodiment of the present invention. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A method for processing form automatic completion, characterized in that: include: Matching similar form data from the form data knowledge base, the similarity of which reaches a specified threshold with the current form subject and the current content identification field in the currently filled-in form information; Generate form completion context information and form completion prompt information respectively according to the currently filled-in form information and the similar form data; The form completion context information and the form completion prompt information are input into a pre-trained Chain Of Thought model based on the Dspy framework to obtain a completion solution for the current remaining form.

2. The method according to claim 1, characterized in that Before matching similar form data whose similarity with the current form subject and the current content identification field in the currently filled-in form information reaches a specified threshold from the form data knowledge base, the method further includes: Use SDK technology to set the basic information of the form and configure the form change monitoring function; wherein the basic information of the form includes the form subject and content identification fields; When the form change monitoring function monitors that the basic form information has changed, the user input behavior data is obtained, and the form filling event formed by the basic form information and the user input behavior data is reported to the form database; wherein the user input behavior data includes the name of the modified field and the filled content; The form filling events in the form database are processed by subject grouping of form data and merging of form data fields to determine final form data, and the final form data is stored in the form knowledge base.

3. The method according to claim 2, characterized in that The subject grouping and field merging processing of the form data on the form filling events in the form database to determine the final form data and storing the final form data in the form knowledge base includes: At regular intervals, incrementally extracting form filling events from the form database; After grouping the form data according to the subject, the last modified field in the form data field is retained to determine the final identification field of the form data; Final form data is determined according to the form data subject and the final identification field of the form data, and the final form data is stored in the form knowledge base.

4. The method according to any one of claims 1 to 3, characterized in that The matching of similar form data from the form data knowledge base, the similarity of which reaches a specified threshold with the current form subject and the current content identification field in the currently filled-in form information, includes: Combine the current form subject and the current content identification field to obtain content to be matched; Calculating the similarity between the processed form data in the form data knowledge base and the content to be matched using a Text Embedding model; The processed form data whose similarity reaches the specified threshold is used as the similar form data.

5. A processing system, characterized in that: The processing system comprises: A matching unit, used to match similar form data from a form data knowledge base whose similarity with the current form subject and the current content identification field in the currently filled-in form information reaches a specified threshold; A generating unit, configured to generate form completion context information and form completion prompt information respectively according to the currently filled-in form information and the similar form data; The input unit is used to input the form completion context information and the form completion prompt information into a pre-trained Chain Of Thought model based on the Dspy framework to obtain a completion solution for the current remaining form.

6. The processing system according to claim 5, characterized in that The processing system comprises: A configuration unit, used for setting the basic form information using SDK technology and configuring the form change monitoring function before the matching unit matches similar form data whose similarity with the current form subject and the current content identification field in the currently filled form information reaches a specified threshold from the form data knowledge base; wherein the basic form information includes the form subject and the content identification field; An acquisition and input unit, used for acquiring user input behavior data when the form change monitoring function monitors that the form basic information has changed, and reporting a form filling event formed by the form basic information and the user input behavior data to a form database; wherein the user input behavior data includes the name of the modified field and the filled content; The processing and storage unit is used to perform form data subject grouping and form data field merging processing on the form filling events in the form database to determine the final form data, and store the final form data in the form knowledge base.

7. The processing system according to claim 6, characterized in that The processing and storage unit comprises: An extraction subunit, used for incrementally extracting form filling events from the form database at intervals of a certain period of time; A retaining subunit is used to retain the last modified field in the form data field after grouping and processing according to the subject of the form data, so as to determine the final identification field of the form data; The determination and storage subunit is used to determine the final form data according to the form data subject and the final identification field of the form data, and store the final form data in the form knowledge base.

8. The processing system according to any one of claims 5 to 7, characterized in that: The matching unit comprises: A combining subunit, used for combining the current form subject and the current content identification field to obtain content to be matched; A calculation subunit, used to calculate the similarity between the processed form data in the form data knowledge base and the content to be matched by using a Text Embedding model; The determination subunit is used to take the processed form data whose similarity reaches the specified threshold as the similar form data.

9. A processing system, characterized in that: The processing system comprises: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the form automatic completion processing method described in any one of claims 1-4.

Images