A software system intelligent interaction method
By pre-training the vertical field big model and user feature model, analyzing and generating task processes, the problems of low interaction efficiency and insufficient intelligent support of traditional software are solved, multilingual and multi-accent interaction are realized, and software usage efficiency and user experience are improved. .
Patent Information
- Application Number
- CN202411001836.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-25
- Publication Date
- 2024-11-12
- Estimated Expiration
- 2044-07-25
AI Technical Summary
Traditional software interaction methods are inefficient in complex application processes and large amounts of data processing. Especially for ordinary users and non-technical users, learning and mastering complex operation processes is a major challenge, and there is a lack of intelligent support for the entire task operation process. , unable to adapt to user needs and habits.
By pre-training the vertical field big model, combining user voice characteristics, usage environment, job responsibilities and usage habit characteristics, a software user interaction feature model is established, user input instructions are analyzed, inference analysis and task flow is generated, and execution scripts are configured to achieve intelligent interaction .
It greatly reduces work intensity, improves software interaction and application efficiency, simplifies operational processes, lowers the learning threshold for users, and allows more non-technical personnel to easily use professional application software.
Smart Images

Figure CN118535713B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of artificial intelligence technology and software application, and in particular to a software system intelligent interaction method. Background Art
[0002] With the rapid development of informatization, digitization and intelligence, various application software systems play an increasingly important role in the design, production, data processing and management decision-making process. Traditional software interaction methods often rely on complex menu navigation and multi-step interactive input operations, which are complicated and inefficient. Especially in application environments that require rapid response and processing of large amounts of data, the limitations of traditional interaction methods are more obvious.
[0003] Existing software interaction methods mainly rely on the graphical user interface (GUI). Users need to use menus, buttons or keyboards to operate the software, and complete data input through input devices such as keyboards, touch screens or writing pens. Although this method is intuitive, its efficiency and convenience are greatly reduced when faced with complex application processes and large amounts of data processing. Especially when faced with complex processes and diverse business needs, users often need to spend a lot of time and energy to familiarize themselves with the various functions and operation processes of the software, and complete data processing and input and output work for a certain task through interactive multi-step combination selection menus. For ordinary users or users with non-technical backgrounds, learning and mastering these complex operation processes is a big challenge.
[0004] Existing software systems have more or less intelligent function support, but they mainly issue simple or single tasks through voice and intelligent analysis and application of processed objects or data, such as question and answer, document analysis, directory generation, content extraction or summary, voice generation of text, text generation of pictures or videos, PPT production, etc., lacking intelligent support for the entire task operation process and unable to make adaptive adjustments based on the actual needs and operating habits of users. This leads to a more complicated software use process and a poor experience, and also limits the performance of software functions, especially for engineering data processing software systems. Summary of the invention
[0005] In view of the above problems, the present invention proposes a software system intelligent interaction method.
[0006] An embodiment of the present invention provides a software system intelligent interaction method, which provides intelligent interaction for software applied in vertical fields. The software system intelligent interaction method includes:
[0007] Based on vertical domain expertise, software function descriptions, human-machine spatial relationships, and typical application task processes, pre-train vertical domain big models to provide understanding and reasoning capabilities for software functions and work content within the domain;
[0008] Based on the user's voice features, usage environment features, job responsibilities features, and usage habit features, a software user interaction feature model is established, wherein the software user interaction feature model enables the user to automatically identify the user during the use of the software and output typical feature parameters of the user under different task processes;
[0009] The user issues a functional task or business requirement instruction to the software system through voice or text input, and the software system analyzes the accent in the user's voice or the user's text input based on the software user interaction feature model to form a valid input instruction;
[0010] Based on the vertical domain big model, the valid input instructions are reasoned and analyzed, and a preset task flow template is called or a new task flow is generated;
[0011] Based on the software user interaction feature model, combined with the user's recent historical operations, and utilizing the vertical domain big model, direct parameters and implicit parameters contained in the effective input instructions are extracted;
[0012] Based on the direct parameters and the implicit parameters, the vertical domain big model is used to understand and reason, the preset task flow template or the new task flow is integrated, and a specific task flow and its execution program script are configured;
[0013] Combined with the application software component interface or API service, the specific task flow and its execution program script are executed at one time, and finally the requirements or tasks proposed by the user are completed.
[0014] Optionally, pre-training the vertical domain big model includes: pre-training or fine-tuning the vertical domain big model of the domain using the data set;
[0015] The data set includes: vertical domain expertise, software function description, and data of typical application task processes, wherein the vertical domain expertise data includes: relevant professional terms, theoretical systems, technical systems, and methods in the field, which are used to understand the expertise input by the user;
[0016] The software function description data includes: software function module structure, function description, function parameter input requirements, and user manual, which is used to interpret user input to the software function module;
[0017] The human-machine spatial relationship includes: the inclusion and association relationship between the person and the spatial area where the software application scenario or data source is located, which is used to identify and understand the spatial and temporal information of spatial objects implicit in the user input;
[0018] Typical application task flow data includes: task flows that use a series of software functions from the perspective of application requirements and business scenarios to convert user input into task flows.
[0019] Optionally, the specific process of establishing the software user characteristic model and outputting the typical characteristic parameters includes:
[0020] Establishing the software user interaction feature model according to the voice feature or the usage environment feature or the job responsibility feature or the usage habit feature;
[0021] Automatically identify the user's identity, output parameters associated with the identity when the user uses the software, and understand and infer the typical characteristic parameters of the user in different task processes;
[0022] The user voice features include: recording the audio feature sequence during the user voice input process, so that the software user feature model supports the recording and recognition of sound features including dialect, tone, and intonation;
[0023] The usage environment features include: the computer environment and spatial association of the user using the software, the computer environment includes the computer hardware, operating system and software environment of the software, and the spatial association includes: user information and its geographical location and spatial affiliation, so that the software user feature model is integrated with the user login information, the user is quickly identified and the user environment and spatial relationship parameters are output;
[0024] The job responsibility characteristics include: according to the corresponding login information, usage role and job responsibility when the user uses the software system, the functions and data within the user's authority range are used as one of the characteristics to identify the user, and at the same time, the implicit parameters required for different task processes are output, and the implicit parameters include the user organization and common limiting conditions required for the operation of the software function;
[0025] The usage habit features include: some features of the user's use of the software, including: usage process, common functions, common data, and software input features. These features are used to identify the user and recommend implicit parameters required for the task process based on the user's input instructions. The implicit parameters include the user's organization and limiting condition input;
[0026] The typical characteristic parameters include: the direct parameters and the implicit parameters.
[0027] Optionally, the text input method includes: describing the function to be executed or the requirement to be completed in a natural language, inputting the described text into the software system, and the software system forms a valid input instruction after a legality check;
[0028] The voice input method includes: describing the function to be executed or the requirement to be completed by voice, analyzing the individual characteristics of the voice, recognizing and converting the voice into text, and then executing the text;
[0029] Among them, both input methods support spoken language, dialects, and personalized terms.
[0030] Optionally, configuring a specific task flow includes:
[0031] Step S1: Obtain the functional task or business requirement instruction input by the user, perform preliminary input cleaning and validity judgment, and confirm that it is a normal user input instruction;
[0032] Step S2: reasoning and analyzing the user input instruction through the vertical domain big model and the software user interaction feature model, and determining and outputting the type of the user input instruction;
[0033] Step S3: According to the type of the user input instruction, if it is a direct software function task, the user input instruction is converted into a specific software function command. If it is a business requirement, combined with the task process reasoning capability of the vertical domain big model, the preset task process template is called, or the user input instruction is converted to generate the new task process, and then the new task process is converted into the corresponding software function command.
[0034] Optionally, extracting direct parameters and implicit parameters contained in the valid input instruction includes:
[0035] Step T1: Based on the software function command or software function task flow, combined with the software function parameter input requirements, the vertical domain big model is used to reason and analyze the user input command to extract direct parameters related to the software function execution;
[0036] Step T2: combining the software function parameter input requirements, comparing the parameters extracted in step T1, analyzing and determining the missing parameters for executing the software function;
[0037] Step T3: According to the missing parameters, combined with the user's recent historical operations and the software user interaction feature model, the understanding and reasoning ability of the vertical field big model is utilized to analyze one by one, and extract the implicit parameters through circular reasoning;
[0038] Step T4: Compare the software function parameter input requirements based on the direct parameters and implicit parameters, and further interact with the user for the parameters that are still missing, or set them through default values.
[0039] Optionally, executing the specific task flow and its execution script at one time to finally complete the requirement or task proposed by the user includes:
[0040] Based on the specific task flow and the direct parameters and the implicit parameters, the task flow is parsed into a plurality of corresponding step scripts that are sequentially executed in the software or call the software to execute according to the order of each step through reasoning and understanding of the vertical domain big model;
[0041] Parse the parameters required in each step script to form a script template for each step of parameter assignment or parameter replacement;
[0042] The direct parameters, the implicit parameters and the script template are integrated to generate a program script executed in steps, and the program script is executed to complete the requirements or tasks proposed by the user.
[0043] Optionally, it also includes:
[0044] If the program script fails to meet the requirements or tasks proposed by the user after execution, the program script is fed back to the user;
[0045] The user partially modifies the program script through user input interaction to satisfy the user's adjustment and control of execution details;
[0046] All program scripts confirmed to be used by the user are saved as a priority reference for subsequent similar tasks, so as to realize task flow fixation, configuration and generation.
[0047] Optionally, the usage environment characteristics further include: human-machine spatial relationship characteristics;
[0048] The human-computer spatial relationship characteristics include: the user's own spatial relationship characteristics and the spatial position when using the software;
[0049] The user's own spatial relationship characteristics include: the implicit spatial association and affiliation of the organization to which the user belongs, which automatically provides the software's task flow with parameters related to the organization structure, data source name, geographic spatiotemporal relationship, business scope, and spatial object name, avoiding repetitive selection and entry.
[0050] Optionally, the specific task process has learning iteration and intelligent recommendation characteristics, and the specific implementation method includes:
[0051] By creating, training, learning and accumulating typical business process samples, the automatic conversion of requirements or instructions to the specific task process is formed, and the accuracy is improved through the learning iteration of the vertical field big model;
[0052] By memorizing the interactive features of users' commonly used business processes, the system automatically recommends the best corresponding task process for users' demand descriptions, spoken expressions, and personalized terminology input by voice or text, and improves the association between the effective input instructions and the specific task process through feedback and iteration, thereby improving the accuracy of intelligent recommendations.
[0053] The software system intelligent interaction method proposed in the present invention, by establishing a vertical field big model for application software, combines the characteristics of user voice, usage environment, job responsibilities, usage habits, etc., to realize the intelligent processing of software interacting in multiple languages, multiple accents or text modes, greatly reducing the work intensity and improving the software interaction and application efficiency. Users can issue task commands to the software system through accented voice, and with the support of the learning, understanding, and reasoning capabilities of the big model, the user's job responsibilities, usage environment, usage habits, business processes and other characteristics are recorded to realize a more intelligent interaction mode, thereby greatly simplifying the operation process, improving work efficiency, and lowering the user's learning threshold at the same time, so that more non-technical personnel can easily use various professional application software systems, with high practicality. BRIEF DESCRIPTION OF THE DRAWINGS
[0054] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:
[0055] Figure 1 It is a flow chart of the software system intelligent interaction method provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0056] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, are only part of the embodiments of the present invention, not all of the embodiments, and are not used to limit the present invention.
[0057] The software system intelligent interaction method proposed in the present invention provides intelligent interaction for software applied in vertical fields. Figure 1 , shows a flow chart of a software system intelligent interaction method of this embodiment, the software system intelligent interaction method comprising:
[0058] Step 101: Based on vertical domain expertise, software function description, human-machine spatial relationship, and typical application task flow, pre-train vertical domain big model to provide understanding and reasoning capabilities of software functions and work content within the domain.
[0059] In order to realize intelligent interaction of software systems, it is first necessary to pre-train a vertical domain big model based on vertical domain expertise, software function description, human-computer spatial relationship and typical application task flow. This vertical domain big model provides the ability to understand and reason about software functions and work content within the domain.
[0060] A better method for pre-training a large vertical domain model includes: using a data set to pre-train or fine-tune a large vertical domain model. Specific data sets include:
[0061] The vertical field expertise, software function description and typical application task process data respectively, among which the vertical field expertise data includes: relevant professional terms, theoretical systems, technical systems and methods in the field, which are used to understand the professional knowledge input by the user.
[0062] The software function description data includes: software function module structure, function description, function parameter input requirements, and user manual, which are used to interpret user input to the software function module.
[0063] The human-computer spatial relationship includes: the inclusion and association relationship between personnel and the spatial area where the software application scenario or data source is located, which is used to identify and understand the spatial and temporal information of spatial objects implicit in user input.
[0064] Typical application task flow data includes: task flows that use a series of software functions from the perspective of application requirements and business scenarios to convert user input into task flows.
[0065] Step 102: Based on the user voice features, usage environment features, job responsibility features, and usage habit features, a software user interaction feature model is established. The software user interaction feature model enables the user to automatically identify the user during the use of the software and output the typical feature parameters of the user under different task processes.
[0066] In addition to the large model of the vertical field, it is also necessary to establish a software user interaction feature model based on user voice features, usage environment features, job responsibilities features, and usage habit features. The software user interaction feature model can automatically identify users during the use of the software and output the typical feature parameters of users under different task processes.
[0067] A better specific process of establishing a software user feature model and outputting the typical feature parameters includes:
[0068] According to the voice features, usage environment features, job responsibilities features, or usage habits features, a software user interaction feature model is established; the user identity is automatically identified, and the parameters associated with the identity when the user uses the software are output, and the typical feature parameters of the user in different task processes are understood and inferred. The so-called typical feature parameters include: the direct parameters and the implicit parameters.
[0069] Among them, user voice features include: recording the audio feature sequence during the user voice input process, so that the software user feature model can support the recording and recognition of sound features including dialects, tone, and intonation.
[0070] The usage environment features include: the computer environment and spatial associations in which the user uses the software. The so-called computer environment includes: the computer hardware, operating system, and software environment in which the software is used; the so-called spatial associations include: user information and its geographical location, spatial affiliation, and spatial associations can enable the software user feature model to be integrated with the user login information, quickly identify the user, and output the user environment and spatial relationship parameters. In addition, the usage environment features also include: human-computer spatial relationship features.
[0071] The so-called human-computer spatial relationship characteristics include: the user's own spatial relationship characteristics and the spatial location when using the software. Among them, the user's own spatial relationship characteristics include: the implicit spatial association and affiliation of the organization to which the user belongs, which automatically provides the software's task flow with relevant parameters for the organization, data source name, geographic spatiotemporal relationship, business scope, and spatial object name, thus avoiding repeated selection and entry.
[0072] Job responsibility characteristics include: based on the corresponding login information, usage role and job responsibilities when the user uses the software system, the functions and data within the user's authority range are used as one of the characteristics to identify the user, and at the same time, the implicit parameters required for different task processes are output. The implicit parameters include the user organization structure and common restriction conditions required for the operation of software functions.
[0073] Usage habit features include: some characteristics of users using the software, which include: usage process, common functions, common data, and software input characteristics. These characteristics are used to identify users and recommend implicit parameters required for task processes based on user input instructions. The implicit parameters include user organization and conditional input.
[0074] Step 103: The user issues a functional task or business requirement instruction to the software system through voice or text input. The software system analyzes the accent in the user's voice or the user's text input based on the software user interaction feature model to form a valid input instruction.
[0075] After the vertical domain big model and the software user interaction feature model are built, users can issue functional tasks or business demand instructions to the software system through voice input or text input during use. The software system can parse the accent in the user's voice or the user's text input based on the software user interaction feature model, thereby forming effective input instructions.
[0076] Preferably, the text input method includes: describing the function to be executed or the requirement to be completed in natural language, inputting the described text into the software system, and the software system forms a valid input instruction after a legality check.
[0077] The voice input method includes: describing the function to be performed or the requirements to be completed by voice, analyzing the individual characteristics of the voice, recognizing and converting the voice into text, and then executing the text. In essence, the so-called voice input method is convenient for users, who can directly input by voice instead of typing. Especially for some special groups of people (such as users with hand injuries or who cannot type), the voice input method allows these special groups to use the software system conveniently. The voice is converted into text after input, which can also be understood as another text input method.
[0078] Regardless of which of the above input methods is used, spoken language, dialects, personalized terms, etc. are supported.
[0079] Step 104: Based on the vertical domain big model, the valid input instructions are reasoned and analyzed, and the preset task flow template is called or a new task flow is generated.
[0080] After obtaining valid input instructions, the valid input instructions can be reasoned and analyzed based on the vertical field big model, and the preset task process template can be called or a new task process can be generated.
[0081] Step 105: Based on the software user interaction feature model, combined with the user's recent historical operations, and using the vertical domain big model, extract the direct parameters and implicit parameters contained in the effective input instructions.
[0082] After obtaining valid input instructions, in order to achieve intelligent interaction, it is also necessary to extract the direct parameters and implicit parameters contained in the valid input instructions based on the software user interaction feature model, combined with the user's recent historical operations, and using the vertical field big model.
[0083] A better method for extracting direct parameters and implicit parameters contained in valid input instructions includes:
[0084] Step T1: Based on the software function command or software function task flow, combined with the software function parameter input requirements, the vertical domain big model is used to reason and analyze the user input command to extract the direct parameters related to the software function execution;
[0085] Step T2: Combine the software function parameter input requirements, compare the parameters extracted in step T1, and analyze and determine the missing parameters for executing the software function;
[0086] Step T3: Based on the missing parameters, combined with the user's recent historical operations and the software user interaction feature model, the understanding and reasoning capabilities of the vertical field big model are used to analyze them one by one and extract the implicit parameters through circular reasoning;
[0087] Step T4: Compare the software function parameter input requirements based on direct parameters and implicit parameters, and further interact with the user for the parameters that are still missing, or set them through default values.
[0088] Step 106: Based on direct parameters and implicit parameters, using the vertical domain big model understanding and reasoning, integrating the preset task flow template or the new task flow, configuring a specific task flow and its execution program script.
[0089] After obtaining the direct parameters and implicit parameters, based on the direct parameters and implicit parameters, the understanding and reasoning of the vertical field big model are used to integrate the preset task process template or new task process to configure a specific task process and its execution program script.
[0090] A better configuration forms a specific task flow including:
[0091] Step S1: Obtain the functional task or business requirement instruction input by the user, perform preliminary input cleaning and validity judgment, and confirm that it is a normal user input instruction;
[0092] Step S2: Perform reasoning analysis on the user input command through the vertical domain big model and the software user interaction feature model, and judge and output the type of the user input command;
[0093] Step S3: According to the type of user input instruction, if it is a direct software function task, the user input instruction is converted into a specific software function command. If it is a business requirement, the task flow reasoning capability of the vertical domain big model is combined to call the preset task flow template, or the user input instruction is converted into a new task flow, and then the new task flow is converted into the corresponding software function command.
[0094] In addition, the specific task process also has the characteristics of learning iteration and intelligent recommendation. The specific implementation methods include:
[0095] By creating, training, learning and accumulating typical business process samples, automatic conversion of requirements or instructions to specific task processes is formed, and the accuracy of configuring specific task processes is improved through learning iterations of large vertical field models.
[0096] By memorizing the interactive features of users' commonly used business processes, the system automatically recommends the best corresponding task process for users' demand descriptions, spoken expressions, and personalized terminology input through voice or text, and improves the correlation between effective input instructions and specific task processes through feedback and iteration, thereby improving the accuracy of intelligently recommended specific task processes.
[0097] Step 107: Combine the application software component interface or API service to execute the specific task flow and its execution program script at one time, and finally complete the needs or tasks proposed by the user.
[0098] After configuring the specific task flow and its execution program script, you can combine the application software component interface or API service to execute the specific task flow and its execution program script at one time, and finally complete the needs or tasks proposed by the user.
[0099] A better method for executing the specific task flow and its execution script at one time to finally complete the requirements or tasks proposed by the user includes:
[0100] First, based on the specific task flow and direct parameters and implicit parameters, the vertical domain large model reasoning and understanding is used to parse the task flow into multiple corresponding step scripts that have a sequence and can be executed in the software or called to execute the software. Generally, a task flow includes several steps, each of which is parsed into a script. These steps are in sequence, so the task flow can be parsed into multiple corresponding step scripts that have a sequence and can be executed in the software or called to execute the software according to the sequence of each step.
[0101] Then, the parameters required in each step script are parsed to form script templates for each step of parameter assignment or parameter replacement; finally, the direct parameters, implicit parameters and script templates are integrated to generate a program script that is executed step by step, and the program script can be executed to complete the user's requirements or tasks.
[0102] In addition, if the program script fails to meet the user's requirements or tasks after execution, the program script needs to be fed back to the user; the user can modify the program script partially through user input interaction to meet the user's adjustment and control of execution details; finally, all program scripts confirmed by the user can be saved as a priority reference for subsequent similar tasks, thereby realizing the fixation, configuration, and generation of task processes.
[0103] In summary, the software system intelligent interaction method proposed in the present invention, by establishing a vertical field large model for application software, combines the characteristics of user voice, usage environment, job responsibilities, usage habits, etc., to realize the intelligent processing of software that interacts in multiple languages, multiple accents or text modes, greatly reducing the work intensity and improving the software interaction and application efficiency. Users can issue task commands to the software system through accented voice, and with the support of the learning, understanding, and reasoning capabilities of the large model, record the user's job responsibilities, usage environment, usage habits, business processes and other characteristics, and realize a more intelligent interaction method, thereby greatly simplifying the operation process and improving work efficiency. At the same time, it reduces the user's learning threshold, so that more non-technical personnel can also easily use various professional application software systems, which has high practicality.
[0104] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.
[0105] Finally, 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 terminal 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 terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.
[0106] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims, all of which are within the protection of the present invention.
Claims
1. A software system intelligent interaction method, characterized in that: The software system intelligent interaction method provides intelligent interaction for software applied in vertical fields, and the software system intelligent interaction method includes: Based on vertical domain expertise, software function description, human-machine spatial relationship and typical application task flow, pre-train vertical domain big model to provide understanding and reasoning ability of software functions and work content in the field, specifically including: using data set to pre-train or fine-tune the vertical domain big model of the field; the data set includes: vertical domain expertise, software function description and typical application task flow data, among which vertical domain expertise data includes: relevant professional terms, theoretical system, technical system and methods in the field, which are used to understand the professional knowledge input by users; software function description data includes: software function module structure, function description, function parameter input requirements, user manual, which are used to interpret user input to software function module; the human-machine spatial relationship includes: the inclusion and association relationship between personnel and the spatial area where the software application scenario or data source is located, which is used to identify and understand the spatial and temporal information of spatial objects implicit in user input; typical application task flow data includes: from the perspective of application requirements and business scenarios, a task flow using a series of software functions is used to convert user input to task flow; Based on the user's voice features, usage environment features, job responsibilities features, and usage habit features, a software user interaction feature model is established. The software user interaction feature model enables the user to automatically identify the user during the use of the software and output the typical feature parameters of the user under different task processes. The usage environment features include: the computer environment and spatial association in which the user uses the software. The computer environment includes the computer hardware, operating system, and software environment in which the software is used. The spatial association includes: user information and its geographic location, and spatial affiliation, so that the software user feature model is integrated with the user login information, the user is quickly identified, and the user environment and spatial relationship parameters are output; The user issues a functional task or business requirement instruction to the software system by voice or text input, and the software system analyzes the accent in the user's voice or the user's text input based on the software user interaction feature model to form a valid input instruction; Based on the vertical domain big model, the valid input instructions are reasoned and analyzed, and a preset task flow template is called or a new task flow is generated; Based on the software user interaction feature model, combined with the user's recent historical operations, the vertical field big model is used to extract the direct parameters and implicit parameters contained in the effective input instructions, specifically including step T1: according to the software function command or software function task flow, combined with the software function parameter input requirements, the vertical field big model is used to reason and analyze the user input instructions, and extract the direct parameters related to the execution of a series of software functions; step T2: combined with the software function parameter input requirements, compared with the parameters extracted in step T1, analyze and determine the missing parameters for executing the software function; step T3: according to the missing parameters, combined with the user's recent historical operations and the software user interaction feature model, the understanding and reasoning ability of the vertical field big model is used to analyze one by one, and extract the implicit parameters through circular reasoning; step T4: according to the direct parameters and implicit parameters, the software function parameter input requirements are compared, and the parameters that are still missing are further interacted with the user, or set by default values; Based on the direct parameters and the implicit parameters, the vertical domain big model is used to understand and reason, and the preset task flow template or the new task flow is integrated to configure and form a specific task flow and its execution program script; Combined with the application software component interface or API service, the specific task flow and its execution program script are executed at one time, and finally the requirements or tasks proposed by the user are completed.
2. The software system intelligent interaction method according to claim 1, characterized in that: The specific process of establishing a software user feature model and outputting the typical feature parameters includes: Establishing the software user interaction feature model according to the voice feature or the usage environment feature or the job responsibility feature or the usage habit feature; Automatically identify the user's identity, output parameters associated with the identity when the user uses the software, and understand and infer the typical characteristic parameters of the user in different task processes; The user voice features include: recording the audio feature sequence during the user voice input process, so that the software user feature model supports the recording and recognition of sound features including dialect, tone, and intonation; The job responsibility characteristics include: according to the corresponding login information, usage role and job responsibility when the user uses the software system, the functions and data within the user's authority range are used as one of the characteristics to identify the user, and at the same time, the implicit parameters required for different task processes are output, and the implicit parameters include the user organization and common limiting conditions required for the operation of the software function; The usage habit features include: some features of the user's use of the software, including: usage process, common functions, common data, and software input features. These features are used to identify the user and recommend implicit parameters required for the task process based on the user's input instructions. The implicit parameters include the user's organization and limiting condition input; The typical characteristic parameters include: the direct parameters and the implicit parameters.
3. The software system intelligent interaction method according to claim 1, characterized in that: The text input method includes: describing the function to be executed or the requirement to be completed in natural language, inputting the described text into the software system, and the software system forms a valid input instruction after passing the legality check; The voice input method includes: describing the function to be executed or the requirement to be completed by voice, analyzing the individual characteristics of the voice, recognizing and converting the voice into text, and then executing the text; Among them, both input methods support spoken language, dialects, and personalized terms.
4. The software system intelligent interaction method according to claim 1, characterized in that: The specific task flow of configuration includes: Step S1: Obtain the functional task or business requirement instruction input by the user, perform preliminary input cleaning and validity judgment, and confirm that it is a normal user input instruction; Step S2: reasoning and analyzing the user input instruction through the vertical domain big model and the software user interaction feature model, and judging and outputting the type of the user input instruction; Step S3: According to the type of the user input instruction, if it is a direct software function task, the user input instruction is converted into a specific software function command. If it is a business requirement, combined with the task process reasoning capability of the vertical domain big model, the preset task process template is called, or the user input instruction is converted to generate the new task process, and then the new task process is converted into a corresponding series of software function commands.
5. The software system intelligent interaction method according to claim 1, characterized in that: Execute the specific task flow and its execution script once and for all to finally complete the requirements or tasks proposed by the user, including: Based on the specific task flow and the direct parameters and the implicit parameters, the task flow is parsed into a plurality of corresponding step scripts that are sequentially executed in the software or call the software to execute according to the order of each step through reasoning and understanding of the vertical domain big model; Parse the parameters required in each step script to form a script template for each step of parameter assignment or parameter replacement; The direct parameters, the implicit parameters and the script template are integrated to generate a program script executed in steps, and the program script is executed to complete the requirements or tasks proposed by the user.
6. The software system intelligent interaction method according to claim 1, characterized in that: Also includes: If the program script fails to meet the requirements or tasks proposed by the user after execution, the program script is fed back to the user; The user partially modifies the program script through user input interaction to satisfy the user's adjustment and control of execution details; All program scripts confirmed to be used by the user are saved as a priority reference for subsequent similar tasks, so as to realize task flow fixation, configuration and generation.
7. The software system intelligent interaction method according to claim 3, characterized in that: The usage environment characteristics also include: human-machine spatial relationship characteristics; The human-computer spatial relationship characteristics include: the user's own spatial relationship characteristics and the spatial position when using the software; The user's own spatial relationship characteristics include: the implicit spatial association and affiliation of the organization to which the user belongs, which automatically provides the software's task flow with parameters related to the organization structure, data source name, geographic spatiotemporal relationship, business scope, and spatial object name, avoiding repetitive selection and entry.
8. The software system intelligent interaction method according to claim 5, characterized in that: The specific task process has the characteristics of learning iteration and intelligent recommendation, and the specific implementation method includes: By creating, training, learning and accumulating typical business process samples, the automatic conversion of requirements or instructions to the specific task process is formed, and the accuracy is improved through the learning iteration of the vertical field big model; By memorizing the interactive features of users' commonly used business processes, the system automatically recommends the best task flow for users' demand descriptions, spoken expressions, and personalized terminology input by voice or text, and improves the association between the effective input instructions and the specific task flow through feedback and iteration to improve the accuracy of intelligent recommendations.
