Intelligent form filling method based on voice interaction

Abstract

The invention discloses an intelligent form filling method based on voice interaction. A form filling system is composed of an intention recognition module, an entity recognition module, a data mapping module, an intelligent error correction module and a dialogue management module. According to the system, the traditional manual form filling operation is replaced by a voice input mode, the complexity and time cost of user input are greatly reduced, and the recognition precision of terminologies is remarkably improved by utilizing a voice recognition and natural language processing model optimized for the financial field, so that the user experience is improved. Specific fields of the form can be accurately recognized according to intention recognition and data mapping functions customized for a specific scene of financial reimbursement, the system is designed based on a multi-round dialogue management module, a user is actively guided to complete complex form filling step by step, the learning cost and the operation difficulty are remarkably reduced, and the user experience is improved. Full-process automation of voice recognition, data mapping and intelligent verification is realized, and manual intervention is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent assisted form filling, and in particular to a method for intelligent form filling based on voice interaction. Background Art

[0002] In the daily operations of modern enterprises and organizations, filling out various forms is one of the most common business processes. With the development of speech recognition technology and natural language processing technology, voice interaction technology has gradually been applied to many fields, especially showing great potential in improving work efficiency and improving user experience. However, existing form filling systems usually still rely on traditional manual input methods or systems based on fixed templates, and fail to fully utilize the advantages of voice interaction technology. Therefore, an intelligent form filling system based on voice intelligent interaction came into being, aiming to make the filling process more intelligent and convenient through speech recognition and natural language processing technology, especially in mobile office and flexible working environments, to provide users with a more efficient and accurate filling experience.

[0003] Although existing technologies have made some improvements in form filling, there are still several major defects and shortcomings:

[0004] Manual input is cumbersome and prone to errors. Currently, most systems rely on users to manually input information, such as invoice amounts, transportation costs, accommodation costs, etc. Since multiple fields and values ​​are involved, manual filling is prone to errors, especially when manually entering numbers, dates or long texts. This not only increases the workload of financial audits, but may also lead to inaccurate data and extended business cycles.

[0005] The process is complicated and not flexible enough

[0006] Existing systems usually require employees to fill out forms in a fixed format, and most systems lack flexible interaction methods. Especially for employees who need to travel frequently and use mobile devices, traditional form filling systems are often not convenient enough. Employees need to manually enter a lot of information on the computer, which is not suitable for modern fast-moving work needs.

[0007] Lack of intelligent processing

[0008] Most existing systems are based on static forms and lack intelligent data processing capabilities. For example, the system is usually unable to intelligently identify and pre-fill commonly used data in forms, and cannot automatically fill in fields based on filling history or preset rules, resulting in users still having to repeatedly enter the same content every time they fill out a form, wasting time and energy.

[0009] Poor user experience and lack of convenient mobile support

[0010] Traditional form systems are mostly designed for desktop computers. For employees who frequently travel, they need to use a computer to fill out forms, which is not convenient for people on business trips. Modern enterprises are increasingly relying on mobile office, and employees often use mobile devices such as mobile phones or tablets to handle expense reimbursement matters. However, existing systems generally lack good mobile support, resulting in users being unable to obtain an efficient expense reimbursement experience when working on the move.

[0011] Lack of smart auditing and compliance verification

[0012] Existing form systems often require manual intervention for review, relying on financial personnel to check and verify the accuracy and compliance of each form, which is a huge workload and inefficient. Especially for a large number of forms, manual review is tedious and prone to errors. The lack of an intelligent review mechanism has greatly reduced the efficiency and accuracy of the reimbursement system.

[0013] Therefore, it is necessary to design an intelligent form filling method based on voice interaction. Summary of the invention

[0014] The purpose of the present invention is to provide a method for intelligently filling in forms based on voice interaction to solve the problems raised in the above-mentioned background technology.

[0015] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0016] The method for intelligent form filling based on voice interaction includes an intention recognition module, an entity recognition module, a data mapping module, an intelligent error correction module and a dialogue management module to form a form filling system, including the following steps:

[0017] S100, guiding the user to interact through the dialogue management module, and converting the voice information into text data;

[0018] S200, using an intention recognition module to identify the operation intention of the converted text data;

[0019] S300, the entity recognition module extracts key fields from the text data and obtains required information fields based on the collected entity report data;

[0020] S400, mapping entity data to specific form fields through a data mapping module;

[0021] S500: Automatically correct and repair the information input by the user through the intelligent error correction module.

[0022] According to the above technical solution, the specific processing method of the dialogue management module in S100 is:

[0023] Design multiple rounds of dialogue in the dialogue management module, and the order of each round of dialogue can be changed. Through multiple rounds of dialogue, the user information that needs to be filled in can be obtained;

[0024] The dialogue management module converts multi-round input dialogue voice information into text data by supporting context memory.

[0025] According to the above technical solution, the specific processing method of the intention recognition module of S200 is:

[0026] Based on natural language understanding technology and intent model, the text data obtained in S100 is identified to obtain the intent category and intent confidence in the text data, and the accurate operation purpose is determined through the intent category information and intent confidence information;

[0027] The intent model is obtained through professional in-depth training in the financial field on the cloud platform.

[0028] According to the above technical solution, the specific processing method of the entity recognition module in S300 is:

[0029] Identify entity reports and obtain the information fields required to fill in different types of reports;

[0030] Extract key fields from text data, including but not limited to reimbursement amount, date, location, transportation, and project name;

[0031] Information fields and key fields are combined with lexical analysis and semantic annotation technology to structure and store user information.

[0032] According to the above technical solution, the specific processing method of the data mapping module in S400 is:

[0033] Through predefined form templates or dynamically generated field mapping rules, structured user information is correctly filled into the corresponding form positions, and at the same time, it is checked whether there are missing fields or format inconsistencies. The field mapping rules can be adjusted based on needs.

[0034] According to the above technical solution, the specific processing method of the intelligent error correction module of S500 is:

[0035] Through logical analysis of recognition results, contextual association checking, and historical data comparison, the system can remind users to make corrections or automatically repair any recognition errors or unreasonable data.

[0036] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0037] 1. Improve form filling efficiency: The system replaces the traditional manual form filling operation with voice input, greatly reducing the complexity and time cost of user input, automatically identifying user intent and core fields, supporting rapid completion of multi-field filling, and reducing repeated interaction steps.

[0038] 2. Enhanced recognition accuracy: The system uses speech recognition and natural language processing models optimized for the financial field to significantly improve the recognition accuracy of professional terms. The system is equipped with an intelligent error correction module that can automatically detect and correct user input errors to ensure the accuracy and completeness of the form content.

[0039] 3. Improve adaptability in the financial field: The intention recognition and data mapping functions customized for specific scenarios of financial reimbursement can accurately identify specific fields of the form, support form templates of multiple financial institutions, have strong compatibility and adaptability, and are suitable for the personalized needs of different units.

[0040] 4. Optimize user interaction experience: The system is designed based on a multi-round dialogue management module, which actively guides users to complete complex form filling step by step, significantly reducing learning costs and operational difficulties.

[0041] 5. Intelligent process processing: realize full process automation of voice recognition, data mapping, and intelligent verification, reduce manual intervention, and in terms of process simplification, automatically omit unnecessary steps through dynamic interactive adjustment mechanism, further improving form filling efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a flowchart of the present invention. DETAILED DESCRIPTION

[0043] The following will be combined with 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 making creative work are within the scope of protection of the present invention.

[0044] like Figure 1 As shown, the present invention provides a technical solution: a method for intelligently filling in a form based on voice interaction, comprising an intention recognition module, an entity recognition module, a data mapping module, an intelligent error correction module and a dialogue management module to form a form filling system, comprising the following steps:

[0045] S100, guiding the user to interact through the dialogue management module, and converting the voice information into text data;

[0046] S200, using an intention recognition module to identify the operation intention of the converted text data;

[0047] S300, the entity recognition module extracts key fields from the text data and obtains required information fields based on the collected entity report data;

[0048] S400, mapping entity data to specific form fields through a data mapping module;

[0049] S500: Automatically correct and repair the information input by the user through the intelligent error correction module.

[0050] Specifically, the specific processing method of the dialogue management module of S100 is:

[0051] Design multiple rounds of dialogue in the dialogue management module, and the order of each round of dialogue can be changed. Through multiple rounds of dialogue, the user information that needs to be filled in can be obtained;

[0052] The dialogue management module converts multi-round input dialogue voice information into text data by supporting context memory.

[0053] Specifically, the specific processing method of the intention recognition module of S200 is:

[0054] Based on natural language understanding technology and intent model, the text data obtained in S100 is identified to obtain the intent category and intent confidence in the text data, and the accurate operation purpose is determined through the intent category information and intent confidence information;

[0055] The intent model is obtained through professional in-depth training in the financial field on the cloud platform.

[0056] Specifically, the specific processing method of the entity recognition module of S300 is:

[0057] Identify entity reports and obtain the information fields required to fill in different types of reports;

[0058] Extract key fields from text data, including but not limited to reimbursement amount, date, location, transportation, and project name;

[0059] Information fields and key fields are combined with lexical analysis and semantic annotation technology to structure and store user information.

[0060] Specifically, the specific processing method of the data mapping module in S400 is:

[0061] Through predefined form templates or dynamically generated field mapping rules, structured user information is correctly filled into the corresponding form positions, and at the same time, it is checked whether there are missing fields or format inconsistencies. The field mapping rules can be adjusted based on needs.

[0062] Specifically, the specific processing method of the intelligent error correction module of S500 is:

[0063] Through logical analysis of recognition results, contextual association checking, and historical data comparison, the system can remind users to make corrections or automatically repair any recognition errors or unreasonable data.

[0064] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0065] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for intelligent form filling based on voice interaction, comprising an intention recognition module, an entity recognition module, a data mapping module, an intelligent error correction module and a dialogue management module to form a form filling system, characterized in that: The steps include: S100, guiding the user to interact through the dialogue management module, and converting the voice information into text data; S200, using an intention recognition module to identify the operation intention of the converted text data; S300, the entity recognition module extracts key fields from the text data and obtains required information fields based on the collected entity report data; S400, mapping entity data to specific form fields through a data mapping module; S500: Automatically correct and repair the information input by the user through the intelligent error correction module.

2. The method for intelligently filling out a form based on voice interaction according to claim 1, characterized in that: The specific processing method of the dialogue management module in S100 is: Design multiple rounds of dialogue in the dialogue management module, and the order of each round of dialogue can be changed. Through multiple rounds of dialogue, the user information that needs to be filled in can be obtained; The dialogue management module converts multi-round input dialogue voice information into text data by supporting context memory.

3. The method for intelligently filling in a form based on voice interaction according to claim 1, characterized in that: The specific processing method of the intention recognition module of S200 is: Based on natural language understanding technology and intent model, the text data obtained in S100 is identified to obtain the intent category and intent confidence in the text data, and the accurate operation purpose is determined through the intent category information and intent confidence information; The intent model is obtained through professional in-depth training in the financial field on the cloud platform.

4. The method for intelligently filling out a form based on voice interaction according to claim 1, characterized in that: The specific processing method of the entity recognition module in S300 is: Identify entity reports and obtain the information fields required to fill in different types of reports; Extract key fields from text data, including but not limited to reimbursement amount, date, location, transportation, and project name; Information fields and key fields are combined with lexical analysis and semantic annotation technology to structure and store user information.

5. The method for intelligently filling in a form based on voice interaction according to claim 1, characterized in that: The specific processing method of the data mapping module of S400 is: Through predefined form templates or dynamically generated field mapping rules, structured user information is correctly filled into the corresponding form positions, and at the same time, it is checked whether there are missing fields or format inconsistencies. The field mapping rules can be adjusted based on needs.

6. The method for intelligently filling in a form based on voice interaction according to claim 1, characterized in that: The specific processing method of the intelligent error correction module of S500 is: Through logical analysis of recognition results, contextual association checking, and historical data comparison, the system can remind users to make corrections or automatically repair any recognition errors or unreasonable data.

Images