Question-and-answer service generation method and apparatus, device, and readable storage medium

Abstract

A question-and-answer service generation method, executed by a computer device and comprising: displaying a function code encapsulation component set in an editing interaction page of a question-and-answer service editing application, wherein the function code encapsulation component set comprises a plurality of function code encapsulation components, and the plurality of function code encapsulation components comprise at least one trigger component, at least one question-and-answer policy component, and at least one language question-and-answer model component (S101); in response to a component selection operation for the function code encapsulation component set, in a question-and-answer service editing page comprised in the editing interaction page, displaying the trigger component, the question-and-answer policy component, and the language question-and-answer model component that are selected by means of the component selection operation, wherein the question-and-answer service editing page is further used for displaying a processing flow connection relationship among the trigger component, the question-and-answer policy component, and the language question-and-answer model component that are selected (S102); and in response to a question-and-answer service confirmation operation triggered in the editing interaction page of the question-and-answer service editing application, on the basis of the processing flow connection relationship among the trigger component, the question-and-answer policy component, and the language question-and-answer model component that are selected, generating a question-and-answer service, wherein the trigger component in the question-and-answer service is used for calling the question-and-answer policy component in the question-and-answer service, the question-and-answer policy component in the question-and-answer service is used for carrying out question-and-answer filtering on a service question when receiving the service question, and the model component in the question-and-answer service is used for outputting a question-and-answer service result on the basis of the filtered service question (S103).

Description

Question and answer service generation method, device, equipment and readable storage medium

[0001] Related applications

[0002] This application claims priority to Chinese patent application number 2023109149947, filed on July 24, 2023, entitled “Question-and-Answer Service Generation Method, Apparatus, Device and Readable Storage Medium,” the entire text of which is incorporated herein by reference. Technical Field

[0003] The present application relates to the field of computer technology, and in particular to a method, apparatus, device, and readable storage medium for generating a question-and-answer service. Background Art

[0004] At present, the technical threshold of large language model question-answering robots is high, and the R&D cost is high. The existing development methods are all manual development based on code. Manual development has low development efficiency and requires a lot of time and energy to complete the task. The implementation threshold is high, and technical personnel are required to have an in-depth understanding of the underlying algorithms of large language models and prepare huge training data sets. The labor cost is high, and manual intervention is required to complete the task. The implementation error rate is high, and human errors and loopholes are prone to occur, which affects the quality and efficiency of the work.

[0005] Summary of the Invention

[0006] According to an embodiment of the present application, a method, apparatus, device, and readable storage medium for generating a question-and-answer service are provided.

[0007] In one aspect, a method for generating a question-answering service is provided, which is executed by a computer device and includes:

[0008] On an editing interaction page of a question-answering service editing application, a function code encapsulation component set is displayed, where the function code encapsulation component set includes multiple function code encapsulation components, and the multiple function code encapsulation components include at least one trigger component, at least one question-answering strategy component, and at least one language question-answering model component;

[0009] In response to a component selection operation on the function code encapsulation component set, displaying the trigger component, question and answer strategy component, and language question and answer model component selected by the component selection operation on a question and answer service editing page included in the editing interaction page, and the question and answer service editing page is further used to display a processing flow connection relationship between the selected trigger component, question and answer strategy component, and language question and answer model component; and

[0010] In response to a question-and-answer service confirmation operation triggered in the editing interaction page of the question-and-answer service editing application, a question-and-answer service is generated based on the processing flow connection relationship between the selected trigger component, the question-and-answer strategy component, and the language question-and-answer model component. The trigger component in the question-and-answer service is used to call the question-and-answer strategy component in the question-and-answer service. The question-and-answer strategy component in the question-and-answer service is used to obtain business questions and then filter the business questions. The model component in the question-and-answer service is used to output question-and-answer service results based on the filtered business questions.

[0011] In another aspect, a device for generating a question-answering service is provided, comprising:

[0012] A component display module, configured to display a function code encapsulation component set on an edit interaction page of a question-and-answer service editing application, wherein the function code encapsulation component set includes multiple function code encapsulation components, each of which includes at least one trigger component, at least one question-and-answer strategy component, and at least one language question-and-answer model component;

[0013] a component editing module for, in response to a component selection operation on the function code encapsulation component set, displaying, on a question-and-answer service editing page included in the editing interaction page, the trigger component, question-and-answer strategy component, and language question-and-answer model component selected by the component selection operation, the question-and-answer service editing page further being configured to display a processing flow connection relationship between the selected trigger component, question-and-answer strategy component, and language question-and-answer model component;

[0014] A component confirmation module is used to respond to a question-and-answer service confirmation operation triggered in the editing interaction page of the question-and-answer service editing application, and generate a question-and-answer service based on the processing flow connection relationship between the selected trigger component, the question-and-answer strategy component, and the language question-and-answer model component. The trigger component in the question-and-answer service is used to call the question-and-answer strategy component in the question-and-answer service. The question-and-answer strategy component in the question-and-answer service is used to filter the business questions after obtaining the business questions. The model component in the question-and-answer service is used to output the question-and-answer service results according to the filtered business questions.

[0015] On the other hand, a computer device is provided, including a processor and a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the processor implements the method provided in the embodiment of the present application.

[0016] On the other hand, a computer-readable storage medium is provided, which stores computer-readable instructions. When the computer-readable instructions are executed by a processor, the processor implements the method provided in the embodiment of the present application.

[0017] On the other hand, a computer program product is provided, comprising computer-readable instructions, which, when executed by a processor, enable the processor to implement the method provided in an embodiment of the present application.

[0018] The details of one or more embodiments of the present application are set forth in the accompanying drawings and the description below. Other features, objects, and advantages of the present application will become apparent from the description, drawings, and claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the conventional technology, the following briefly introduces the drawings required for use in the embodiments or the conventional technology descriptions. Obviously, the drawings described below are merely embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the disclosed drawings without any creative work.

[0020] FIG1 is a schematic diagram of a network architecture provided by an embodiment of the present application;

[0021] FIG2 is a schematic diagram of a scenario for generating a question-and-answer service provided in an embodiment of the present application;

[0022] FIG3 is a schematic diagram of another scenario of a question-and-answer generation service provided in an embodiment of the present application;

[0023] FIG4 is a flow chart of a method for generating a question-and-answer service according to an embodiment of the present application;

[0024] FIG5 is a schematic diagram of a scenario for generating a question-and-answer service provided in an embodiment of the present application;

[0025] FIG6 is a schematic diagram of a process for accessing a question-and-answer service in a communication application according to an embodiment of the present application;

[0026] FIG7 is a flow chart of a method for generating a question-and-answer service according to an embodiment of the present application;

[0027] FIG8 is a schematic diagram of the structure of a device for generating a question-and-answer service provided in an embodiment of the present application;

[0028] FIG9 is a schematic structural diagram of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0029] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0030] Artificial intelligence (AI) technology is a comprehensive discipline encompassing a wide range of fields, encompassing both hardware and software technologies. Foundational AI technologies generally include sensors, specialized AI chips, cloud computing, distributed storage, big data processing, operating / interaction systems, and mechatronics. AI software technologies primarily encompass computer vision, speech processing, natural language processing, and machine learning / deep learning.

[0031] Deep learning (DL) is the process of learning the inherent patterns and representational hierarchies of sample data. The information gained during this learning process is highly helpful in interpreting data such as text, images, and sounds. Its ultimate goal is to enable machines to acquire the same analytical and learning capabilities as humans, enabling them to recognize data such as text, images, and sounds. Deep learning is a complex machine learning algorithm that has achieved results in speech and image recognition that far surpass previous technologies.

[0032] The solutions provided in the embodiments of this application involve artificial intelligence, which will be specifically described through the following embodiments.

[0033] Please refer to Figure 1, which is a schematic diagram of a network architecture provided by an embodiment of the present application. The network architecture schematic diagram may include a work server 100, a terminal device 200, a communication server 300, and a terminal device 400, wherein the work server 100 may have a communication connection with the terminal device 200, the work server 100 may have a communication connection with the communication server 300, and the communication server 300 may have a communication connection with the terminal device 400. The above-mentioned communication connection does not limit the connection method, and may be directly or indirectly connected through a wired communication method, directly or indirectly connected through a wireless communication method, or through other methods, which are not limited by the present application.

[0034] It should be understood that the terminal device 200 can be installed with a question-and-answer service editing application. When the question-and-answer service editing application runs in the terminal device 200, it can interact with the work server 100 for data. The work server 100 is also the background server of the question-and-answer service editing application. The terminal device 200 can generate a question-and-answer service through the question-and-answer service editing application, and the question-and-answer service can be accessed through the communication application. The terminal device 400 can be installed with a communication application. When the communication application runs in the terminal device 400, it can interact with the communication server 300 for data. The communication server 300 is also the background server of the communication application. Among them, the question-and-answer service refers to an intelligent service that can generate answers to questions through a language question-and-answer model. The question-and-answer service can be an independent client or an embedded sub-client integrated in the communication application. The work server 100 and the communication server 300 can be integrated into the same server, and the terminal device 200 and the terminal device 400 can also be the same terminal device, which is not limited here.

[0035] Specifically, the terminal device 200 is installed with a question-and-answer service editing application. The question-and-answer service editing application of the terminal device 200 can display a function code encapsulation component set, which includes one or more trigger components, one or more question-and-answer strategy components, and one or more language question-and-answer model components.

[0036] Among them, the function code encapsulation component refers to a pre-compiled function component with a preset code function. The function code encapsulation component can be a component that is encapsulated using at least one of a pre-written program module or library and has a predetermined function function, which is used to perform function processing on the business questions input by the user. The trigger component is a component that triggers the workflow of the question and answer service. It is the starting component for the operation of the entire question and answer service workflow. The trigger component can be used to detect whether the business question raised by the user meets the question and answer trigger conditions, and can call the question and answer strategy component in the question and answer service when the question meets the question and answer trigger conditions. The question and answer strategy component is a component that filters business questions and can filter business questions that the language question and answer model component cannot handle. The language question and answer model component can perform semantic recognition on the filtered business questions and output the question and answer service results of the business questions.

[0037] The terminal device 200 can trigger a component selection operation in the question-and-answer service editing application, and select a function code encapsulation component in the question-and-answer service editing page of the question-and-answer service editing application, wherein the function code encapsulation component can be selected by dragging the function code encapsulation component displayed in the question-and-answer service editing page, or by creating a new function code encapsulation component. The selected function code encapsulation component is displayed on the editing page of the question-and-answer service editing application, and then a processing flow connection relationship is created between the selected function code encapsulation components by connecting lines. The processing flow connection relationship can represent the process sequence of at least two function code encapsulation components. The connection can be automatic connection given a preset rule, or it can be a connection triggered in response to a connection operation on the function code encapsulation component.

[0038] The terminal device 200 can trigger the component configuration operation in the question-and-answer service editing application, display the parameter configuration page, and further modify the component information of the function code encapsulation component in the parameter configuration page of the question-and-answer service editing application. The component information may include the processing flow connection relationship, the instance name, and the instance template identifier. The instance name is the name of the customized function code encapsulation component. The instance template identifier is used to determine the function function applied by the function code encapsulation component, and the business problem is processed by the function function. For example, the permission judgment subcomponent in the question-and-answer strategy component can use the function function userlist.find() to search in the user table uesrlist (question permission range) that has the permission to ask questions, and filter out the business problems that do not return results.

[0039] The terminal device 200 can generate a question-and-answer service through the processing flow connection relationship between the selected trigger component, the question-and-answer strategy component, and the language question-and-answer model component in response to the trigger component confirmation operation in the question-and-answer service editing application.

[0040] The question-and-answer service can be accessed by a communication application. The terminal device 400 is installed with a communication application. The communication application of the terminal device 400 can display a question-and-answer interaction page, and the user can enter business questions on the question-and-answer interaction page. Among them, the business question includes a trigger parameter. When the trigger parameter meets the question-and-answer trigger condition indicated by the trigger component in the question-and-answer service, the terminal device 400 can send the business question to the communication server 300. The communication server 300 then forwards the business question to the work server 100. The work server 100 processes the business question through the function code encapsulation component selected by the question-and-answer service, obtains the question-and-answer service result of the business question, and returns the question-and-answer service result to the communication server 300. The communication server 300 sends the question-and-answer service result to the terminal device 400. The communication application in the terminal device 400 can display the question-and-answer service result of the business question entered by the user on the question-and-answer interaction page. The question-and-answer interaction page can be a window, a floating layer or other form, which is not limited here.

[0041] The embodiments of the present application can provide various types of function code packaging components that support independent deployment, realize flexible access of function code packaging components, and generate question and answer services through the processing flow connection relationship between the function code packaging components selected by the user. It allows people who do not have professional programming knowledge to develop question and answer services through visual function code packaging components and processing flow connection relationships, rather than writing code to develop question and answer services, thereby improving the production efficiency and versatility of question and answer service development, and reducing development costs and implementation error rates.

[0042] Please refer to Figure 2, which is a schematic diagram of a scenario for generating a question and answer service provided by an embodiment of the present application. A question and answer service editing application can be installed on the terminal device 200. In the editing interaction page 210 of the question and answer service editing application, the question and answer service editing application can display a function code encapsulation component set in the component candidate page 212. The code encapsulation component set can include a trigger component (such as an application trigger component and a form trigger component), a question and answer strategy component (such as a question judgment sub-component and a frequency control sub-component) and multiple language question and answer model components of different algorithm types.

[0043] The user can click a selection control 211 to trigger a component selection operation in the Q&A service editing application. The user can then drag and drop the selected function code package component from the component candidate page 212 to the Q&A service editing page 213 of the Q&A service editing application. The component candidate page displays the function code package components for selection. The selected function code package component can be displayed and edited in the Q&A service editing page.

[0044] The Q&A service editing application in the terminal device 200 may display the selected trigger component, Q&A strategy component, and language Q&A model component. For example, the terminal device 200 may display the trigger component (including the application trigger component), the Q&A strategy component (including the question determination subcomponent, the frequency control subcomponent, and the authority determination subcomponent), and the language Q&A model component in the editing interaction page 220 of the Q&A service editing application.

[0045] The application trigger component can be the starting component for the entire Q&A service workflow. Once the Q&A service editing application is completed, the Q&A service is established. When a user accesses the Q&A service, they enter a business question on the independent page of the Q&A service. Terminal device 400 can then activate the functions provided by the Q&A service through the application trigger component and call the Q&A strategy component within the Q&A service.

[0046] The question judgment subcomponent in the Q&A strategy component is used to filter business questions that contain prohibited business terms or whose file formats are in the question handling exception format. Prohibited business terms are words that are not allowed in business questions. They can be illegal or illegal words, or words that are composed solely of symbols, spaces, etc. and cannot express the actual linguistic meaning. The question handling exception format is specifically set for the selected language question answering model component. For example, if the selected language question answering model component can only process text messages, the question handling exception format can be set to a file format for business questions such as images and audio that the language question answering model cannot process. The frequency control subcomponent in the Q&A strategy component is used to filter business questions where the frequency of the same user calling the Q&A service exceeds the filtering threshold indicated by the frequency control subcomponent. The unit time range for the frequency of calling the Q&A service can be one hour, one day, or one week, and there is no restriction here. The permission judgment subcomponent in the Q&A strategy component is used to filter business questions where the user's question permission does not fall within the question permission scope.

[0047] The editing interaction page 220 of the question-and-answer service editing application includes a control display area, in which a connection control 221 is displayed. When a user clicks the connection control 221, the terminal device 200 can trigger a component arrangement operation in response to the click of the connection control 221. In response to the component arrangement operation, the terminal device 200 drags the connection control 221 from the application trigger component to the problem judgment subcomponent, creates a processing flow connection relationship between the application trigger component and the problem judgment subcomponent, and determines the problem judgment subcomponent as the function code encapsulation component of the next process of the application trigger component. Similarly, at the problem judgment subcomponent, the user can use the connection control 221 to determine that the frequency control subcomponent is the function code encapsulation component of the next process of the problem judgment subcomponent. Similarly, the permission judgment subcomponent can be determined as the function code encapsulation component of the next process of the frequency control subcomponent, and the language question-and-answer model component can be determined as the function code encapsulation component of the next process of the permission judgment subcomponent, thereby completing the creation of the processing flow connection relationship of the selected function code encapsulation component.

[0048] The control display area of ​​the edit interaction page 220 of the question-and-answer service editing application may also display a configuration control 222. When the user clicks the configuration control 222, the terminal device 200 may trigger a component configuration operation in response to the click on the configuration control 222. After triggering the component configuration operation, the terminal device 200 may, in response to a click on the question judgment subcomponent, display a parameter configuration page 223 in the question-and-answer service editing application to configure the business banned words and question handling exception format of the selected question judgment subcomponent. Configuring business banned words and question handling exception formats may, for example, set business banned words to banned word group A and determine image format and audio format as question handling exception formats.

[0049] The control display area of ​​the editing interaction page 220 of the question-and-answer service editing application can also display a save control 224. When the user clicks the save control 224, the terminal device 200 can trigger the question-and-answer service confirmation operation in response to the click of the save control 224. In response to the question-and-answer service confirmation operation, the terminal device 200 can generate a question-and-answer service by giving the user the processing flow connection relationship between the selected trigger component, question-and-answer strategy component, and language question-and-answer model component through the question-and-answer service editing application.

[0050] It is understandable that in the specific implementation of this application, the user data involved (such as business issues, question frequency and question authority, etc.) needs to obtain user permission or consent when the above and following embodiments of this application are applied to specific products or technologies, and the collection, use and processing of relevant data need to comply with the relevant laws, regulations and standards of relevant countries and regions.

[0051] Please refer to Figure 3, which is a schematic diagram of a scenario for generating a question-and-answer service provided by an embodiment of the present application. A communication application may be installed on the terminal device 400, and the terminal device 400 may access the question-and-answer service through the communication application. In the question-and-answer interaction page 410 of the question-and-answer service editing application, the terminal device 400 can display the question-and-answer interaction page in the communication application. The user can enter a business question A in the question-and-answer interaction page. The business question A includes a trigger parameter. If the trigger parameter meets the question-and-answer trigger condition indicated by the trigger component in the question-and-answer service (which can be when the user enters a business question on an independent page of the question-and-answer service, the question-and-answer service is called), the business question A is sent to the communication server 300 through the trigger component. The communication server 300 forwards the business question A to the work server 100. The work server 100 processes the business question A through the function code encapsulation component selected by the question-and-answer service, obtains the question-and-answer service result of the business question A, and returns the question-and-answer service result to the communication server 300. The communication server 300 sends the question-and-answer service result to the terminal device 400. As shown in the question-and-answer interaction page 420 of the question-and-answer service editing application, the communication application in the terminal device 400 can display the question-and-answer service result A of the business question A entered by the user in the question-and-answer interaction page.

[0052] Please refer to Figure 4, which is a flow chart of a method for generating a question-and-answer service provided in an embodiment of the present application. The method for generating a question-and-answer service can be executed by a computer device, which can be the terminal device 200 or the work server 100 shown in Figure 1. As shown in Figure 4, the method may include:

[0053] Step S101: Displaying a function code encapsulation component set on an editing interaction page of a question-answering service editing application, where the function code encapsulation component set includes multiple function code encapsulation components, each of which includes at least one trigger component, at least one question-answering strategy component, and at least one language question-answering model component.

[0054] The computer device may display a function code encapsulation component set in a question and answer service editing application, where the function code encapsulation component set includes one or more trigger components, one or more question and answer strategy components, and one or more language question and answer model components.

[0055] Specifically, the computer device can display a function code encapsulation component set in the editing interaction page of the question-and-answer service editing application, and the function code encapsulation set can include multiple function code encapsulation components. The multiple function code encapsulation components include at least one trigger component, at least one question-and-answer strategy component, and at least one language question-and-answer model component. The function code encapsulation component set can also include other function code encapsulation components, such as function code encapsulation components for message notifications, function code encapsulation components for calling API interfaces of other platforms, function code encapsulation components for terminating processes, and function code encapsulation components for performing empty operations, which are not limited in the embodiments of the present application. Empty operations are operations that do not perform actual processing. The computer program must know what operations need to be performed. When no operation is required, it can be represented by empty operations.

[0056] Among them, the function code encapsulation component refers to a component with function functions and predefined behaviors and properties. The properties can be the code logic and name, parameters, return value, etc. inside the function code encapsulation component. The function function can be filled into the function code encapsulation component by using various pre-written program modules and libraries. The trigger component can be used to detect whether the business question raised by the user meets the question and answer trigger conditions, and can call the question and answer strategy component in the question and answer service when the question meets the question and answer trigger conditions. The question and answer strategy component is a component that filters business questions and can filter business questions that the language question and answer model component cannot handle. The language question and answer model component can perform semantic recognition on the filtered business questions and output the question and answer service results of the business questions.

[0057] Different trigger components indicate different problem trigger conditions. For example, trigger components may include application trigger components, form trigger components, mention trigger components, entry trigger components, and periodic trigger components. Among them, the problem trigger condition of the application trigger component may be that the user calls the Q&A service by entering a business question on the single chat page. The problem trigger condition of the form trigger component may be that the user submits a form to call the Q&A service. The problem trigger condition of the mention trigger component may be that the user mentions (can be @) the Q&A service (or a specific name representing the Q&A service) in the chat to call the Q&A service. The problem trigger condition of the entry trigger component may be that the user switches from the chat page to the single chat page of the Q&A service to call the Q&A service. The problem trigger condition of the periodic trigger component may not require the user to actively call it, but may be automatically called by a time period, work (event) cycle, etc. For example, the Q&A strategy component in the Q&A service may be periodically called every day. The problem trigger condition of the trigger component can be determined by different functional functions and specific workflows, and the embodiments of the present application are not limited here.

[0058] Different Q&A strategy components filter business Q&A in different ways. For example, a Q&A strategy component can include a question determination subcomponent, a frequency control subcomponent, and a permission determination subcomponent. The question determination subcomponent can filter business questions that contain prohibited business terms or whose file formats are in the question handling exception format. Prohibited business terms refer to words that are not allowed in business questions. These can be illegal or non-compliant words, or words that are composed solely of symbols, spaces, and other elements and cannot express the actual linguistic meaning. The question handling exception format is specifically set for the selected language Q&A model component. For example, if the selected language Q&A model component can only process text messages, the question handling exception format can be set to a file format for business questions such as images and audio that the language Q&A model cannot process. The frequency control subcomponent is used to filter business questions whose frequency of calling the Q&A service by the same user exceeds the filtering threshold indicated by the frequency control subcomponent. The unit time range for the frequency of calling the Q&A service can be one hour, one day, or one week, and there is no restriction here. The permission determination subcomponent is used to filter business questions whose question permissions do not fall within the scope of the user's question permissions.

[0059] It can be understood that the question-and-answer strategy component can filter business questions through various logical judgments and direct the process in different directions. For example, business questions in the permission judgment subcomponent where the user's question permission falls within the question permission scope can flow to the language question-and-answer model component, while business questions where the user's question permission does not fall within the question permission scope can flow to the function code encapsulation component that performs no operations and the function code encapsulation component for the message notification of "no question permission".

[0060] The language question-answering models built in different language question-answering model components use different algorithm types. The algorithm type can refer to the training text sets of the language question-answering models coming from different vertical fields, such as e-commerce customer service, communication chat assistant, and intelligent education question-answering. Alternatively, the algorithm type can refer to the different initial language models used and the different training methods, such as recurrent neural networks (RNNs) and long short-term memory networks (LSTMs). This is not a limitation in the present embodiment.

[0061] Among them, the training process of the language Q&A model established in the language Q&A model component can be as follows: Obtain a training text set, filter out business stop words in the training text set, perform text segmentation on the filtered training text set to obtain M text segments; count the occurrence times of each of the M text segments respectively to obtain the text frequencies corresponding to the M text segments, where M is a positive integer; obtain an initial language model, establish a text dictionary of the initial language model based on the M text segments and the M text frequencies, and the text dictionary includes word vectors corresponding to the M text segments respectively; obtain Q&A samples, and construct Q&A word vectors corresponding to the Q&A samples through the text dictionary; train the initial language model through the Q&A word vectors to obtain the language Q&A model component.

[0062] Specifically, the training process of the language Q&A model in the language Q&A model component can be as follows: Obtain a training text set, filter out business stop words in the training text set, perform text segmentation on the filtered training text set to obtain M text segments, count the occurrence times of the M text segments to obtain the text frequencies corresponding to the M text segments respectively; M is a positive integer; obtain an initial language model, establish a text dictionary of the initial language model based on the M text segments and the M text frequencies; the text dictionary includes word vectors corresponding to the M text segments respectively; obtain Q&A samples, construct Q&A word vectors corresponding to the Q&A samples through the text dictionary, and adjust the model parameters of the initial language model through the Q&A word vectors until the initial language model meets the model convergence condition, then obtain the language Q&A model component.

[0063] Specifically, a computer device can obtain a training text set. The training text set can include conventional natural language materials (such as web texts, news texts, novel texts, etc.), and can also include technical corpora for business needs (such as educational texts, etc.). The computer device can perform data cleaning on the training text set, re-examine and verify the training text set, and delete duplicate, missing-value, and invalid-value training texts to obtain a training text set with data consistency.

[0064] The computer device can remove business stop words from the training text set after data cleaning. Business stop words refer to words or punctuation marks that are common in natural language texts but have no specific meaning in semantic analysis. For example, in Chinese texts, business stop words can include: "de", "le", "shi", "zai", "you", etc. In English texts, business stop words can include: "the", "a", "an", "in", etc.

[0065] A computer device can perform text segmentation (also known as word segmentation) on a training text set after removing business stop words. Text segmentation can be done by rule matching or probability matching. Taking the text segmentation of training text C in training text set A by probability matching as an example, there will be a word segmentation scheme S for the text segmentation of training text C. The task of text segmentation is to find the scheme with the highest probability P(S|C) of the segmented words in training text set A. According to Bayes' formula, the probability P(S|C) can be calculated as shown in formula (1):

[0066] Among them, P(C) is the probability of training text C appearing in training text set A, and P(C|S) is the probability of restoring training text C by the word segmentation scheme. The word segmentation scheme S divides the training text set into M text segments (text segments W1, text segments W2,..., text segments W i , i , i , ,

[0067] , i ,

[0070] , i ,

[0069] , ,

[0068] , , ), then P(S) = ∑[[ID=,7]] M＝i P(W i ),P(W i ) is the ratio of the number of occurrences of text segment W i in training text set A to the total number of words in training text set A.

[0067] The computer device can count the number of occurrences of text segment W i in training text set A to obtain the text frequencies corresponding to the M text segments respectively. The text frequencies can also be the probabilities of the text segments in training text set A, that is, P(W i ).

[0068] The computer device can establish a text dictionary of the initial language model through the M text segments and the M text frequencies. Among them, the text dictionary can include word vectors corresponding to the M text segments respectively. The word vector is a numerical representation form that converts text data into a form acceptable to the model. It can also include the position of text segment W i in training text C, the number of text segment W i , and the restored word form of text segment W i . Restoring the word form means restoring the text segment to its original form considering the context and word nature of the text segment. For example, both "ate dinner" and "had dinner" are restored to "eat dinner".

[0069] The initial language model can be a neural network model, such as a recurrent neural network or a long short-term memory network. Among them, the initial language model can be composed of multiple hidden layers and an output layer.

[0070] The computer device can set the hidden layer size and learning rate of the initial language model. The learning rate refers to the gradient descent step size determined based on the model's convergence conditions, which affects the speed of model convergence. The gradient refers to the adjustment of the hidden layer parameter values ​​in the initial language model during the gradient descent algorithm.

[0071] The computer device can convert the word vectors in the text dictionary into question-answering training sequences and question-answering target sequences of the initial language model. The question-answering training sequences and question-answering target sequences are also question-answering samples.

[0072] The computer device can input the question-answer training sequence into the initial language model and train it according to a predefined loss function to obtain the question-answer prediction result output by the initial language model at the output layer. The loss function can be a cross entropy loss function, a mean squared error loss function, etc., without limitation.

[0073] The computer device can adjust the parameter values ​​in the hidden layer of the initial language model based on the calculated gradient to minimize the loss function. When the question-answering prediction results of the initial language model and the question-answering target sequence meet a preset model convergence condition, a language question-answering model component is obtained. The model convergence condition refers to the performance metric between the question-answering prediction results and the question-answering target sequence being lower than a preset value. The performance metric may include accuracy, precision, recall, etc.

[0074] It is understandable that during the training of the initial language model, techniques such as dropout (a hidden layer in a neural network model that randomly sets some neurons to 0 during training to reduce the risk of overfitting of the neural network model) and batch normalization (which can normalize the input of each small batch so that the mean of each feature in the training data is close to 0 and the variance is close to 1, thereby accelerating the training process and improving the stability and generalization ability of the neural network model) can be used to improve the training effect of the model.

[0075] Step S102, in response to the component selection operation for the function code encapsulation component set, the trigger component, question and answer strategy component and language question and answer model component selected by the component selection operation are displayed in the question and answer service editing page included in the editing interaction page. The question and answer service editing page is also used to display the processing flow connection relationship between the selected trigger component, question and answer strategy component and language question and answer model component.

[0076] Specifically, please refer to Figure 5, which is a scenario diagram of generating a question and answer service provided in an embodiment of the present application. As shown in Figure 5, the question and answer service editing application can mainly include components such as a flow orchestration canvas, an application selector, an image rendering engine, an application parameter configurator, a debugger and a version controller.

[0077] The flow orchestration canvas can be used to display the functional code encapsulation components. The application selector can be used to locate the functional code encapsulation components on the Q&A service editing page when dragging them, ensuring that the user selects the correct functional code encapsulation component. The graphics rendering engine can graphically render the functional code encapsulation components and the connections between them (processing flow connection relationships). Circular modules can represent trigger components; diamond modules mainly represent Q&A strategy components; and square modules represent specific application actions, such as sending messages, calling API interfaces of other platforms, terminating processes, and no operations. The application parameter configurator can graphically configure functional code components and enter relevant configuration information in the application parameter configurator, such as the instance template identifier described below. The debugger can run the Q&A service and generate the execution log table described below, so that the running status and service quality quantitative data of the Q&A service can be counted through log information. The version controller can save different versions of the Q&A service. The version controller can be used to switch between different versions of the Q&A service and record information about each version change.

[0078] The computer device can respond to the component selection operation of the question and answer service editing application on the function code encapsulation component set (which can be through the application selector shown in Figure 5), and drag the selected function code encapsulation component to the question and answer service editing page of the question and answer service editing application. It can also first select the required function code encapsulation component and directly create the selected function code encapsulation component at the specified location on the question and answer service editing page of the question and answer service editing application.

[0079] The computer device can respond to the component orchestration operation of the question-and-answer service editing application, and create a processing flow connection relationship between the functional code encapsulation components through wiring. For example, it can create a processing flow connection relationship between the trigger component and the question-and-answer strategy component. At the trigger component, it can connect to the question-and-answer strategy component to determine the question-and-answer strategy component as the functional code encapsulation component of the next process of the trigger component.

[0080] In one embodiment, the method for generating a question and answer service further includes: in response to a first component orchestration operation on a function code encapsulation component selected in a question and answer service editing page, creating a processing flow connection relationship between a selected trigger component and a question and answer strategy component; based on the processing flow connection relationship between the selected trigger component and the question and answer strategy component, determining the selected trigger component as a function code encapsulation component for executing a starting process of the question and answer service, and determining the selected question and answer strategy component as a function code encapsulation component for executing a next process of the trigger component; in response to a second component orchestration operation on a function code encapsulation component selected in a question and answer service editing page, creating a processing flow connection relationship between the selected question and answer strategy component and a language question and answer model component; based on the processing flow connection relationship between the selected question and answer strategy component and the language question and answer model component, determining the selected language question and answer model component as a function code encapsulation component for executing the next process of the question and answer strategy component.

[0081] Optionally, the computer device may respond to a first component orchestration operation between the trigger component and the question and answer strategy component and a second component orchestration operation between the question and answer strategy component and the language question and answer model component. The orchestration process may be: responding to the first component orchestration operation of the question and answer service editing application on the function code encapsulation component selected in the question and answer service editing page, creating a processing flow connection relationship between the selected trigger component and the question and answer strategy component, and based on the processing flow connection relationship between the selected trigger component and the question and answer strategy component, determining the selected trigger component as the function code encapsulation component for executing the starting process of the question and answer service, and determining the selected question and answer strategy component as the function code encapsulation component for executing the next process of the trigger component; responding to the second component orchestration operation of the question and answer service editing application on the function code encapsulation component selected in the question and answer service editing page, creating a processing flow connection relationship between the selected question and answer strategy component and the language question and answer model component, and based on the processing flow connection relationship between the selected question and answer strategy component and the language question and answer model component, determining the selected language question and answer model component as the function code encapsulation component for executing the next process of the question and answer strategy component.

[0082] Specifically, the computer device can respond to the question and answer service editing application's first component orchestration operation between the selected trigger component and the question and answer strategy component in the question and answer service editing page, and can connect to the question and answer strategy component at the trigger component to create a processing flow connection relationship between the selected trigger component and the question and answer strategy component, determine the selected trigger component as a functional code encapsulation component for executing the question and answer service starting process, and determine the selected question and answer strategy component as a functional code encapsulation component for executing the next process of the trigger component.

[0083] Similarly, the computer device can also respond to the question-and-answer service editing application's second component orchestration operation on the question-and-answer service editing page, between the selected question-and-answer strategy component and the language question-and-answer model component, by connecting the question-and-answer strategy component to the language question-and-answer model component, creating a processing flow connection relationship between the selected question-and-answer strategy component and the language question-and-answer model component, and determining the selected language question-and-answer model component as a functional code encapsulation component for executing the next flow of the question-and-answer strategy component. The question-and-answer strategy component can include a question determination subcomponent, a frequency control subcomponent, and an authority determination subcomponent.

[0084] In one embodiment, the question-and-answer strategy component includes a question judgment subcomponent, a frequency control subcomponent, and a permission judgment subcomponent. The method for generating a question-and-answer service also includes: in response to a third component orchestration operation on the question judgment subcomponent and the frequency control subcomponent selected in the question-and-answer service editing page, creating a processing flow connection relationship between the selected question judgment subcomponent and the frequency control subcomponent; based on the processing flow connection relationship between the selected question judgment subcomponent and the frequency control subcomponent, determining the selected question judgment subcomponent as a function code encapsulation component for executing the question-and-answer strategy starting process, and determining the selected frequency control subcomponent as a function code encapsulation component for executing the next process of the question judgment subcomponent; in response to a fourth component orchestration operation on the frequency control subcomponent and the permission judgment subcomponent selected in the question-and-answer service editing page, creating a processing flow connection relationship between the selected frequency control subcomponent and the permission judgment subcomponent; based on the processing flow connection relationship between the selected frequency control subcomponent and the permission judgment subcomponent, determining the selected permission judgment subcomponent as a function code encapsulation component for executing the next process of the frequency control subcomponent.

[0085] Optionally, the process of the computer device arranging the problem judgment subcomponent, the frequency control subcomponent and the authority judgment subcomponent can be: responding to the third component orchestration operation of the question judgment subcomponent and the frequency control subcomponent selected in the question and answer service editing page by the question and answer service editing application, creating a processing flow connection relationship between the selected problem judgment subcomponent and the frequency control subcomponent, and based on the processing flow connection relationship between the selected problem judgment subcomponent and the frequency control subcomponent, determining the selected problem judgment subcomponent as a functional code encapsulation component for executing the question and answer strategy starting process, and determining the selected frequency control subcomponent as a functional code encapsulation component for executing the next process of the problem judgment subcomponent; responding to the fourth component orchestration operation of the question and answer service editing application on the frequency control subcomponent and the authority judgment subcomponent selected in the question and answer service editing page, creating a processing flow connection relationship between the selected frequency control subcomponent and the authority judgment subcomponent, and based on the processing flow connection relationship between the selected frequency control subcomponent and the authority judgment subcomponent, determining the selected authority judgment subcomponent as a functional code encapsulation component for executing the next process of the frequency control subcomponent.

[0086] Specifically, the computer device can respond to the question and answer service editing application's third component arrangement operation between the selected question judgment subcomponent and the frequency control subcomponent in the question and answer service editing page, and can connect to the frequency control subcomponent at the question judgment subcomponent to create a processing flow connection relationship between the selected question judgment subcomponent and the frequency control subcomponent, determine the selected question judgment subcomponent as a functional code encapsulation component for executing the question and answer strategy starting process, and determine the selected frequency control subcomponent as a functional code encapsulation component for executing the next process of the question judgment subcomponent.

[0087] Similarly, the computer device can respond to the question and answer service editing application's fourth component arrangement operation between the selected frequency control sub-component and the authority judgment sub-component in the question and answer service editing page, and can connect to the frequency control sub-component at the question judgment sub-component to create a processing flow connection relationship between the selected frequency control sub-component and the authority judgment sub-component, and determine the selected authority judgment sub-component as a functional code encapsulation component for executing the next process of the frequency control sub-component.

[0088] The question-and-answer service editing application in the computer device can display the selected trigger component, question-and-answer strategy component, and language question-and-answer model component on the question-and-answer service editing page, and display the processing flow connection relationship between the selected trigger component, question-and-answer strategy component, and language question-and-answer model component.

[0089] In one embodiment, the method for generating a question-and-answer service also includes: in response to a component configuration operation for a function code encapsulation component selected in a question-and-answer service editing page, displaying multiple instance template identifiers in a parameter configuration page in an editing interaction page; determining a target instance template identifier among the multiple instance template identifiers; in the question-and-answer service editing page, updating the selected function code encapsulation component based on the target instance template identifier, the instance template identifier is used to determine the function function corresponding to the function code encapsulation component, and the function function is used to perform functional processing on business problems.

[0090] Optionally, the computer device may respond to a component configuration operation on a function code encapsulation component, and the configuration process may be: responding to a component configuration operation of a question and answer service editing application on a function code encapsulation component selected in a question and answer service editing page, displaying multiple instance template identifiers in a parameter configuration page of the question and answer service editing application, and determining a target instance template identifier among the multiple instance template identifiers; the multiple instance template identifiers include a target instance template identifier; in the question and answer service editing page of the question and answer service editing application, updating the selected function code encapsulation component based on the target instance template identifier; the instance template identifier is used to determine the function function corresponding to the function code encapsulation component, and the function function is used to perform function processing on business problems.

[0091] Specifically, the computer device can respond to a component configuration operation on the function code encapsulation component (perhaps through the application parameter configurator shown in FIG5 ) by displaying multiple instance template identifiers on the parameter configuration page of the Q&A service editing application. The instance template identifiers can be used to identify the function function corresponding to the function code encapsulation component, and the function code encapsulation component uses the function function to perform functional processing on the business question. The computer device can select a desired instance template identifier from the multiple instance template identifiers and update the function code encapsulation component using the selected instance template identifier. For example, in the periodic trigger component, select the time trigger template identifier or the event trigger template identifier. For example, in the periodic trigger, select the time trigger template identifier and modify the instance parameter to 86400 seconds, i.e., 1 day. Then, the Q&A service will be triggered periodically every 1 day. Select the event trigger template identifier and modify the instance parameter (event) to "temperature exceeding 36°C." Then, the Q&A service will be automatically triggered when the temperature exceeds 36°C.

[0092] Optionally, the computer device may also configure the instance name of the function code encapsulation component, the instance name of the previous process node, the instance name of the next process node, and instance parameters on the parameter configuration page of the Q&A service editing application. The instance name and the instance name of the next process node may collectively represent the processing flow connection relationship, and the instance parameters may be auxiliary parameters of the function corresponding to the instance template identifier.

[0093] The computer device can modify the processing flow connection relationship between the function code encapsulation component and the function code encapsulation component of the previous process node or the next process node by modifying the instance name of the previous process node or the next process node. The computer device can modify the instance parameters. For example, in the periodic trigger, select the time trigger template identifier and modify the instance parameter to 86400 seconds, i.e., 1 day. Then, the Q&A service will be triggered periodically every other day.

[0094] The computer device can update the function code package component through the selected instance template identifier (or the determined instance name, the instance name of the selected previous process node, the instance name of the selected next process node, the determined instance parameters, etc.) in the question and answer service editing page of the question and answer service editing application.

[0095] In one embodiment, the question-and-answer service generation method further includes: in response to a first component configuration operation on the question judgment subcomponent selected in the question-and-answer service editing page, configuring the business banned words and question handling exception format of the selected question judgment subcomponent in the parameter configuration page in the editing interaction page; in response to a second component configuration operation on the frequency control subcomponent selected in the question-and-answer service editing page, configuring the frequency threshold of the selected frequency control subcomponent in the parameter configuration page; in response to a third component configuration operation on the permission judgment subcomponent selected in the question-and-answer service editing page, configuring the question permission range of the selected permission judgment subcomponent in the parameter configuration page.

[0096] Optionally, the computer device may respond to a first component configuration operation of the question judgment subcomponent, a second component configuration operation of the frequency control subcomponent, and a third component configuration operation of the authority judgment subcomponent in the question and answer strategy component. The configuration process may be: responding to the first component configuration operation of the question judgment subcomponent selected in the question and answer service editing page by the question and answer service editing application, configuring the business banned words and question handling exception format of the selected question judgment subcomponent in the parameter configuration page of the question and answer service editing application; responding to the second component configuration operation of the frequency control subcomponent selected in the question and answer service editing page by the question and answer service editing application, configuring the frequency threshold of the selected frequency control subcomponent in the parameter configuration page; responding to the third component configuration operation of the authority judgment subcomponent selected in the question and answer service editing page by the question and answer service editing application, configuring the question authority range of the selected authority judgment subcomponent in the parameter configuration page.

[0097] Specifically, the computer device can respond to the first component configuration operation of the question judgment subcomponent selected in the question and answer service editing page of the question and answer service editing application, and configure the business banned words and question handling exception format of the selected question judgment subcomponent in the parameter configuration page of the question and answer service editing application.

[0098] Because question-and-answer services require relatively high computing costs, computer devices can choose to use a question judgment subcomponent to perform a layer of validity judgment on the business questions entered by users, filtering out some typical invalid questions. For example, business questions that contain banned business words or whose file formats are abnormal question processing formats can be filtered out. The filtered business questions are introduced into the empty operation without any processing.

[0099] The question handling exception format is specifically configured for the selected language question answering model component. For example, if the selected language question answering model component can only process text messages, the question handling exception format can be set to a file format such as images or audio that the language question answering model cannot handle. Business banned words can refer to words that are not allowed in business questions, can be illegal or unauthorized, or can be custom words.

[0100] Optionally, the problem judgment subcomponent can also filter business problems that only contain trigger parameters (for example, in the cut-in trigger component, the user cuts into the single chat page but does not enter a business problem), and other business problems without business problem content (for example, empty messages that lack business problem content during transmission). Business problems without actual natural semantics, such as those that only consist of symbols, emoticons (such as emoji / s, etc.), spaces, special characters (such as the color value #66ccff, the markup language

[0101]

[0102]

[0103]

[0104]

[0105]

[0106]

[0107]

[0108]

[0109]

[0110]

[0111]

[0112]

[0113]

[0114]

[0115]

[0116]

[0117]

[0118]

[0119]

[0120]

[0121]

[0122]

[0123]

[0124]

[0125]

[0126]

[0127]

[0128]

[0129]

[0130]

[0131]

[0132]

[0133]

[0134]

[0135]

[0136]

[0137]

[0138]

[0139]

[0140]

[0141]

[0142]

[0143]

[0144]

[0145]

[0146]

[0147]

[0148]

[0149]

[0150]

[0151]

[0152]

[0153]

[0154]

[0155]

[0156]

[0157]

[0158]

[0159]

[0160]

[0161]

[0162]

[0163]

[0164]

[0165]

[0166]

[0167]

[0168]

[0169]

[0170]

[0171]

[0172]

[0173]

[0174]

[0175]

[0176]

[0177]

[0178]

[0179]

[0180]

[0181]

[0182]

[0183]

[0184]

[0185]

[0186]

[0187]

[0188]

[0189]

[0190]

[0191]

[0192]

[0193]

[0194]

[0195]

[0196]

[0197]

[0198]

[0199]

[0200]

[0201]

[0202]

[0203]

[0204]

[0205]

[0206]

[0207]

[0208]

[0209]

[0210]

[0211]

[0212]

[0213]

[0214]

[0215]

[0216]

[0217]

[0218]

[0219]

[0220]

[0221]

[0222]

[0223]

[0224]

[0225]

[0226]

[0227] The computer device can respond to the Q&A service editing application's second component configuration operation on the frequency control subcomponent selected on the Q&A service editing page. On the parameter configuration page, the computer device can configure the frequency threshold of the selected frequency control subcomponent to control the user's question frequency (the frequency of invoking the Q&A service). Business questions whose question frequency by the same user exceeds the filtering threshold set by the frequency control subcomponent are filtered out. The unit time range for the question frequency can be one hour, one day, or one week, without limitation. For example, the unit time range can be set to 86,400 seconds, i.e., one day. The number of questions is set to 100. This means that the same user is allowed to call the Q&A service 100 times per day. After the user calls the Q&A service more than 100 times, subsequent business questions will be filtered out. The computer device can respond to the Q&A service editing application's third component configuration operation on the permission judgment subcomponent selected on the Q&A service editing page. On the parameter configuration page, the computer device can configure the question permission range of the selected permission judgment subcomponent to filter out business questions whose question permission does not fall within the question permission range. The user calling the Q&A service can be compared with a built-in user list. If the user who calls the Q&A service is not in the built-in user list, the business question of the user who calls the Q&A service is answered, and a reminder that the user has no permission to ask questions is returned to the user. Step S103, in response to the Q&A service confirmation operation triggered in the editing interaction page of the Q&A service editing application, a Q&A service is generated based on the processing flow connection relationship between the selected trigger component, the Q&A strategy component, and the language Q&A model component. The trigger component in the Q&A service is used to call the Q&A strategy component in the Q&A service. The Q&A strategy component in the Q&A service is used to filter the business questions after obtaining the business questions. The model component in the Q&A service is used to output the Q&A service results based on the filtered business questions. Specifically, the computer device can respond to the Q&A service confirmation operation of the Q&A service editing application and generate a Q&A service through the processing flow connection relationship between the selected trigger component, the Q&A strategy component, and the language Q&A model component. A question-and-answer service is a service that generates answers to questions through a language question-and-answer model. The generated question-and-answer service can be directly accessed and used in applications supported by the question-and-answer service editing application. For example, if the question-and-answer service editing application supports directly embedding the question-and-answer service in a communication application, the question-and-answer service can be accessed in the communication application.In one embodiment, the method for generating the question-and-answer service further includes: creating an execution log table for the question-and-answer service based on the selected trigger component, question-and-answer strategy component, and language question-and-answer model component; recording log information of the function code encapsulation component in the question-and-answer service through the execution log table; collecting quantitative data on the operating status and service quality of the question-and-answer service based on the log information; and adjusting at least one of the trigger component, question-and-answer strategy component, language question-and-answer model component, or processing flow connection relationship in the question-and-answer service based on the operating status and service quality quantitative data of the question-and-answer service. Optionally, the computer device can display the log information of the question and answer service during its operation, and adjust the question and answer service through the log information. The process of displaying the log information of the question and answer service during its operation can be: based on the selected trigger component, question and answer strategy component and language question and answer model component, create an execution log table for the question and answer service, and record the log information of the functional code encapsulation component in the question and answer service through the execution log table; display the log information of the question and answer service during its operation through the execution log table, collect quantitative data on the operation status and service quality of the question and answer service based on the log information, and adjust at least one of the trigger component, question and answer strategy component, language question and answer model component or processing flow connection relationship in the question and answer service based on the operation status and service quality quantitative data of the question and answer service. Specifically, the computer device can create an execution log table for the question-and-answer service using the selected trigger component, question-and-answer strategy component, and language question-and-answer model component. The execution log table can record log information for the function code encapsulation component in the question-and-answer service. The execution log table can also include log levels and log content classifications. The log levels can be as shown in Table 1: Table 1. The log content classifications can be as shown in Table 2: Table 2. The computer device can display log information during the operation of the question-and-answer service using the execution log table, and use the log information to calculate the operating status and service quality quantitative data of the question-and-answer service. For example, if a user configures the periodic trigger component of the question-and-answer service and updates the selected time trigger (Timer Trigger) template identifier to an event trigger (Event Trigger) template identifier, but does not configure the triggered event (i.e., the instance parameter corresponding to the event trigger template identifier), resulting in the periodic trigger component being unable to call the question-and-answer service. The execution log table can then record the log information as having a log level of ERROR, a log content classification of configuration log, and a detailed record of the user's modification behavior. The service quality quantitative data can include the occurrence rate of error log levels, the error rate of question-and-answer service results, etc., which are not limited here. Computer equipment can use the running status and service quality quantitative data of the question-answering service to adjust the trigger components, question-answering strategy components, language question-answering model components, processing flow connection relationships between the above components, instance template identifiers, and instance parameters in the question-answering service, thereby reducing work errors and manual intervention and improving the accuracy and consistency of the question-answering service.For example, if the running status of the question-and-answer service shows that the load is too high, the frequency threshold of the frequency control subcomponent in the question-and-answer strategy component can be adjusted to increase the frequency threshold to reduce the running load of the question-and-answer service and improve the service quality quantitative data of the question-and-answer service. It can be understood that the embodiment of the present application can provide various types of function code encapsulation components that support independent deployment, realize flexible access of function code encapsulation components, and generate question-and-answer services through the processing flow connection relationship between the function code encapsulation components selected by the user. The question-and-answer service generally includes three categories: trigger components, question-and-answer strategy components, and language question-and-answer model components, rather than using fixed function code encapsulation components and the processing flow relationship of function code encapsulation components. Users can flexibly select the required function code encapsulation components according to business needs and adjust the processing flow connection relationship therein to generate a question-and-answer service that meets business needs. The embodiment of the present application displays a function code encapsulation component set in a question and answer service editing application; the function code encapsulation component set includes at least one trigger component, at least one question and answer strategy component and at least one language question and answer model component; in response to the component selection operation of the question and answer service editing application for the function code encapsulation component set, the selected trigger component, question and answer strategy component and language question and answer model component are displayed in the question and answer service editing page of the question and answer service editing application; the question and answer service editing page is also used to display the processing flow connection relationship between the selected trigger component, question and answer strategy component and language question and answer model component; in response to the question and answer service confirmation operation in the question and answer service editing application, a question and answer service is generated based on the processing flow connection relationship between the selected trigger component, question and answer strategy component and language question and answer model component; the trigger component in the question and answer service is used to call the question and answer service, the question and answer strategy component in the question and answer service is used to filter business questions, and the model component in the question and answer service is used to output question and answer service results based on the filtered business questions. It can provide various functional code packaging components that support independent deployment, realize flexible access of functional code packaging components, and generate question-and-answer services through the processing flow connection relationship between the functional code packaging components selected by the user. It allows people who do not have professional programming knowledge to develop question-and-answer services through visual functional code packaging components and processing flow connection relationships, rather than writing code to develop question-and-answer services, thereby improving the production efficiency and versatility of question-and-answer service development and reducing development costs and implementation error rates.In one embodiment, the process of accessing the Q&A service in a communication application may specifically be: displaying the Q&A interaction page of the Q&A service in the communication application, obtaining the business question entered in the Q&A interaction page, the business question including the trigger parameter; if the trigger parameter satisfies the Q&A trigger condition indicated by the trigger component, calling the Q&A strategy component in the Q&A service through the trigger component; filtering the business questions through the Q&A strategy component in the Q&A service to obtain the filtered business questions; performing Q&A recognition on the filtered business questions through the language Q&A model component in the Q&A service, and outputting the Q&A service result; displaying the Q&A service result for the business question in the Q&A interaction page. In the embodiment of the present application, a pre-edited function code encapsulation component is used to implement access to the Q&A service in the communication application, which is efficient, convenient, has low development cost, and low implementation error rate. In one embodiment, the method for generating a question and answer service further includes: obtaining an execution record table for the question and answer service, the execution record table including an instance template identifier; performing question and answer recognition on filtered business questions through a language question and answer model component in the question and answer service, and outputting a question and answer service result, including: determining the filtered business questions as output data variables of a trigger component, and determining the output data variables of the trigger component as input data variables of a question and answer strategy component; determining a function function corresponding to the question and answer strategy component based on the instance template identifier, performing function processing on the input data variables of the question and answer strategy component through the function function corresponding to the question and answer strategy component, to obtain output data variables of the question and answer strategy component; determining the output data variables of the question and answer strategy component as input data variables of the language question and answer model component; determining a function function corresponding to the language question and answer model component based on the instance template identifier; performing function processing on the input data variables of the language question and answer model component through the function function corresponding to the language question and answer model component, to obtain output data variables of the language question and answer model component; and determining the output data variables of the language question and answer model component as the question and answer service result. Please refer to Figure 6, which is a flowchart of accessing the question-and-answer service in a communication application provided by an embodiment of the present application. Accessing the question-and-answer service in a communication application is implemented by a computer device, which may be the terminal device 400 or the communication server 300 shown in Figure 1. As shown in Figure 6, the method may include: step S201, obtaining an execution record table for the question-and-answer service, the execution record table containing an instance template identifier. Specifically, after the question-and-answer service is edited, the computer device can create an execution record table for the question-and-answer service through the component information in the trigger component, question-and-answer strategy component, and language question-and-answer model component selected by the question-and-answer service, that is, save the DSL Json structure represented by the function code encapsulation component to the background database. When the trigger component calls the question-and-answer strategy component, the computer device can read the execution record table to complete the execution of the question-and-answer service.Component information may include the instance name, the instance name of the next process node, instance parameters, the instance name of the previous process node, and the instance template identifier. The instance name and the instance name of the next process node, or the instance name of the previous process node, may represent the connection relationship between the processing flow. The question-and-answer strategy component may include a question determination subcomponent, a frequency control subcomponent, and a permission determination subcomponent. DSL Json is a fast, lightweight JSON parsing library. It supports basic JSON types such as numbers, strings, arrays, and objects. For example, the execution record table may record the periodic trigger component as shown in Table 3: Table 3. A computer device can obtain the input data variables and output data variables of the function function corresponding to the function code encapsulation component using the instance name in the execution record table. When calling the question-and-answer service, the input data variables and output data variables can be replaced with corresponding data (such as business questions) using the execution record table. The input data variables refer to the input of the function code encapsulation component in the processing flow, and the output data variables refer to the output of the function code encapsulation component in the processing flow. Step S202: Display a Q&A interactive page of the Q&A service in the communication application, retrieve a business question entered in the Q&A interactive page, and obtain the business question. The business question includes trigger parameters. If the trigger parameters meet the Q&A trigger conditions indicated by the trigger component, the trigger component invokes the Q&A strategy component in the Q&A service. A computer device may display a Q&A interactive page in the communication application, and enter a business question. The business question includes trigger parameters. If the trigger parameters meet the Q&A trigger conditions indicated by the trigger component, the trigger component invokes the Q&A strategy component in the Q&A service. Specifically, the computer device may display a Q&A interactive page in the communication application, and the user enters the business question in the Q&A interactive page. The business question may include trigger parameters, which are parameters included by the user when sending the business question. The business question may also include prompts. Prompts can be words or sentences designed based on user needs and application scenarios. For example, the business question "What to eat today" may include prompts such as "lunch," "Chinese food," and "takeout." The prompt word is the restored word form (including contextual information and semantic information) used during the training of the initial language model in step S101 of the embodiment of Figure 4. If the trigger parameter meets the question-answering trigger condition indicated by the trigger component in the question-answering service, the computer device can activate the function provided by the question-answering service through the trigger component and call the question-answering strategy component in the question-answering service. Among them, the question-answering service refers to a service that generates answers to questions through a language question-answering model.Optionally, the computer device can encrypt the business question of the question and answer strategy component in the call question and answer service to start the initial process of the question and answer service. The process can be: displaying the question and answer interaction page of the question and answer service in the communication application, obtaining the business question entered in the question and answer interaction page, and the business question includes a trigger parameter; if the trigger parameter meets the question and answer trigger condition indicated by the trigger component, the message credential is obtained through the trigger component, and the business question and the message credential are sent to the communication server, so that the communication server encrypts the business question according to the message credential to obtain the encrypted ciphertext, and sends the encrypted ciphertext to the working server. The message credential is also stored in the working server. The working server is used to decrypt the encrypted ciphertext through the stored message credential to obtain the business question. The working server is also used to output the question and answer service result corresponding to the business question through the question and answer service. Specifically, the computer device can encrypt the business question of the question-and-answer strategy component in the call question-and-answer service and start the initial process of the question-and-answer service. The process can be: displaying a question-and-answer interaction page in the communication application, entering the business question in the question-and-answer interaction page; the business question includes trigger parameters; if the trigger parameters meet the question-and-answer trigger conditions indicated by the trigger component, the trigger component sends the business question and the message certificate to the communication server corresponding to the communication application, so that the communication server encrypts the business question according to the message certificate to obtain an encrypted ciphertext, and sends the encrypted ciphertext to the work server corresponding to the question-and-answer service editing application; the message certificate is also stored on the work server; the work server is used to decrypt the encrypted ciphertext using the message certificate to obtain the business question, and the work server is also used to output the question-and-answer service result corresponding to the business question through the question-and-answer service. Specifically, the work server can encapsulate the question-and-answer service and call the question-and-answer service by providing an external interface, such as constructing a network request (HTTP request). The computer device can write the business question in the request body (Body part) of the network request message, and the work server returns the final question-and-answer service result to the computer device via a message (HTTP Response). The computer device can establish a Q&A service in the communication application and obtain a message credential for the Q&A service, where the message credential may include the callback address (webhook) of the work server, an encryption token (Token), and a communication key (Adaptive Communication Environment key, ACE key). The computer device can set the callback address of the Q&A service to the callback address of the work server in the communication application. The computer device can display a Q&A interaction page in the communication application, where the user can enter a business question. The business question may include trigger parameters, which are parameters carried by the user when sending the business question.If the trigger parameters meet the Q&A trigger conditions indicated by the trigger component in the Q&A service, the computer device can send the business question and message credentials to the communication server corresponding to the communication application through the trigger component. The communication server can encrypt the received business question using the encryption token and communication key to obtain encrypted ciphertext, and the communication server can send the encrypted ciphertext to the working server. The working server can receive the encrypted ciphertext through the trigger component, decrypt the encrypted ciphertext using the encryption token and communication key stored in the working server, obtain the business question, and invoke the Q&A strategy component in the Q&A service. The working server can filter the business question using the Q&A strategy component in the Q&A service, perform Q&A recognition on the filtered business question using the language Q&A model component in the Q&A service, output the Q&A service result, send the Q&A service result to the computer device, and display the Q&A service result on the Q&A interaction page in the communication application. Step S203: The filtered business question is determined as the output data variable of the trigger component, and the output data variable of the trigger component is determined as the input data variable of the Q&A strategy component. Step S204: Determine the function corresponding to the question-and-answer strategy component based on the instance template identifier. Perform functional processing on the input data variables of the question-and-answer strategy component using the function corresponding to the question-and-answer strategy component to obtain the output data variables of the question-and-answer strategy component. The output data variables of the question-and-answer strategy component are determined as the input data variables of the language question-and-answer model component. The computer device can determine the function corresponding to the question-and-answer strategy component based on the instance template identifier in the execution record table. Perform functional processing on the input data variables of the question-and-answer strategy component using the function corresponding to the question-and-answer strategy component to determine the output data variables of the question-and-answer strategy component as the input data variables of the language question-and-answer model component. Specifically, the computer device can trigger the component to use the received business question as the input data variable of the question-and-answer strategy component. The computer device can determine the function corresponding to the question-and-answer strategy component using the instance template identifier in the execution record table. For example, the permission judgment subcomponent in the question-and-answer strategy component can use the function userlist.find() to search the table uesrlist (question permission range) of users with question permission and filter out business questions that do not return results. The computer device can perform functional processing on the input data variables of the question-answering strategy component through the functional function corresponding to the question-answering strategy component to obtain the output data variables of the question-answering strategy component (i.e., the business question filtered by the question-answering strategy component). The computer device can determine the output data variables of the question-answering strategy component as the input data variables of the language question-answering model component.In one embodiment, a question-answering strategy component in a question-answering service includes a question judgment subcomponent, a frequency control subcomponent, and an authority judgment subcomponent. Business questions are filtered through the question-answering strategy component in the question-answering service to obtain filtered business questions, including: obtaining question-answering statistics of users who initiate business questions, the question-answering statistics including question frequency and question authority, the question frequency being the frequency of users calling the question-answering service, and the question authority being determined based on the user authority of the user; obtaining business questions, filtering the business questions through business banned words and question handling exception formats indicated by the question judgment subcomponent, and transmitting the business questions filtered by the question judgment subcomponent to the frequency control subcomponent; filtering the business questions filtered by the question judgment subcomponent through the frequency threshold and question frequency indicated by the frequency control subcomponent, and transmitting the business questions filtered by the frequency control subcomponent to the authority judgment subcomponent; filtering the business questions filtered by the frequency control subcomponent through the question authority range and question authority indicated by the authority judgment subcomponent to obtain filtered business questions. The process of filtering business questions by a computer device through the question judgment subcomponent, the frequency control subcomponent, and the authority judgment subcomponent in the question-answering strategy component can be as follows: obtaining the question-answering statistical information of the user who initiated the business question; the question-answering statistical information includes the question frequency and the question authority, the question frequency is the frequency of the user calling the question-answering service; the question authority is determined based on the user's user authority; obtaining the business question, filtering the business question through the business banned words and the question handling exception format indicated by the question judgment subcomponent, and transmitting the business question filtered by the question judgment subcomponent to the frequency control subcomponent; filtering the business question filtered by the question judgment subcomponent through the frequency threshold and the question frequency indicated by the frequency control subcomponent, and transmitting the business question filtered by the frequency control subcomponent to the authority judgment subcomponent; filtering the business question filtered by the frequency control subcomponent through the question authority range and the question authority indicated by the authority judgment subcomponent, and transmitting the business question filtered by the authority judgment subcomponent to the language question-answering model component in the question-answering service. Specifically, the computer device can obtain the question-answering statistical information of the user who initiated the business question. Among them, question-and-answer statistics include question frequency and question permissions. Question frequency is the frequency with which users call the question-and-answer service, and question permissions are determined by the user's user permissions. The computer device can obtain the business questions transmitted by the trigger component (i.e., the output data variable of the trigger component), and filter the business questions using the business prohibited words and question handling exception formats indicated by the question judgment subcomponent. Business questions containing business prohibited words or whose file formats are in question handling exception formats are filtered out. Business prohibited words are words that are not allowed to be used in business questions. They can be illegal words or words that are composed solely of symbols, spaces, etc. and cannot express actual linguistic meaning.The question handling exception format is specifically set for the selected language question-answering model component. For example, if the selected language question-answering model component can only process text messages, the question handling exception format can be set to a file format for business questions such as images and audio that the language question-answering model cannot process. The computer device can transmit the business questions filtered by the question determination subcomponent to the frequency control subcomponent. The frequency control subcomponent indicates the frequency threshold and question frequency, and filters the business questions filtered by the question determination subcomponent. The frequency control subcomponent filters out business questions whose frequency of invoking the question-answering service by the same user exceeds the filtering threshold indicated by the frequency control subcomponent. The unit time range of the frequency of invoking the question-answering service can be one hour, one day, or one week, without limitation. The computer device can transmit the business questions filtered by the frequency control subcomponent to the permission determination subcomponent. The permission determination subcomponent indicates the question permission range and question permission, and filters out business questions whose question permission does not fall within the question permission range. The question permission can also be determined by the user's user permissions. User permissions refer to objects such as administrators or chat group members in communication applications. The computer device can transmit the business questions filtered by the authority judgment subcomponent to the language question and answer model component in the question and answer service (the business questions filtered by the authority judgment subcomponent are determined as the input data variables of the language question and answer model component). The embodiment of the present application filters some business questions by the question and answer strategy component (which may include the question judgment subcomponent, the frequency control subcomponent and the authority judgment subcomponent, etc.), which can reduce the waste of computing costs of the language question and answer model and improve the quality and production efficiency of the question and answer service. Step S205, based on the instance template identifier, determine the function function corresponding to the language question and answer model component; through the function function corresponding to the language question and answer model component, perform function processing on the input data variables of the language question and answer model component to obtain the output data variables of the language question and answer model component. The computer device can determine the function function corresponding to the language question and answer model component based on the instance template identifier in the execution record table, perform function processing on the input data variables of the language question and answer model component through the function function corresponding to the language question and answer model component, and determine the output data variables of the language question and answer model component as the question and answer service result. Specifically, the computer device can determine the functional function corresponding to the language question and answer model component through the instance template identifier in the execution record table, and perform functional processing on the input data variables of the language question and answer model component (business questions after filtering by the authority judgment sub-component) through the functional function corresponding to the language question and answer model component.In one embodiment, the language question answering model component in the question answering service is used to perform question answering recognition on the filtered business questions and output the question answering service results, including: removing business stop words in the filtered business questions to obtain updated business questions; performing text segmentation on the updated business questions to obtain N text segments, where N is a positive integer; converting the N text segments into text word vectors respectively, encoding the text word vectors to obtain context information and semantic information of the text word vectors; and performing semantic recognition on the context information and semantic information through the language question answering model component to obtain question answering service results for the business questions. The language question-answering model component can first perform data preprocessing (including filtering business stop words and text segmentation) on the filtered business questions by the authority judgment subcomponent, and convert them into text word vectors, and then perform question-answering recognition. The process can be: obtaining the business questions filtered by the question-answering strategy component, removing the business stop words in the filtered business questions, obtaining updated business questions, performing text segmentation on the updated business questions, and obtaining N text fragments; N is a positive integer; converting the N text fragments into text word vectors respectively, encoding the text word vectors, obtaining the context information and semantic information of the text word vectors, and performing semantic recognition on the context information and semantic information based on the language question-answering model component to obtain the question-answering service results for the business questions. Specifically, the process in which the computer device can perform data preprocessing (including filtering business stop words and text segmentation) on the business questions filtered by the authority judgment subcomponent and convert them into text word vectors can be seen in the specific description of step S101 of the embodiment of Figure 4, which will not be repeated here. The computer device can encode the text word vectors to obtain the context information and semantic information of the text word vectors. The context information and semantic information of the text word vector are the restored word forms of the text fragments. The computer device can perform semantic recognition on the context information and semantic information through the language question and answer model component to obtain the question and answer service results of the business problem. After the question and answer service is processed, the execution record table will also store the input data variables and output data variables of each function code encapsulation component. Step S206, display the question and answer service results for the business problem on the question and answer interaction page. Specifically, the computer device can display the question and answer service results for the business problem on the question and answer interaction page in the communication application.In the embodiment of the present application, a question-and-answer interaction page is displayed in a communication application, and a business question is input into the question-and-answer interaction page; the business question includes a trigger parameter; if the trigger parameter meets the question-and-answer trigger condition indicated by the trigger component, the question-and-answer strategy component in the question-and-answer service is called through the trigger component; based on the selected trigger component, the question-and-answer strategy component and the language question-and-answer model component, an execution record table for the question-and-answer service is created, and the component information of the function code encapsulation component in the question-and-answer service is recorded through the execution record table; the component information includes a processing flow connection relationship, an instance name and an instance template identifier for characterizing the function code encapsulation component in the question-and-answer service; the business question is determined as an output data variable of the trigger component, and the output data variable of the trigger component is determined as a question-and-answer strategy component according to the processing flow connection relationship indicated by the execution record table. The function corresponding to the question-answering strategy component is determined based on the instance template identifier in the execution record table, and the input data variable of the question-answering strategy component is processed by the function corresponding to the question-answering strategy component to obtain the output data variable of the question-answering strategy component, and the output data variable of the question-answering strategy component is determined as the input data variable of the language question-answering model component; the function corresponding to the language question-answering model component is determined based on the instance template identifier in the execution record table, and the input data variable of the language question-answering model component is processed by the function corresponding to the language question-answering model component to obtain the output data variable of the language question-answering model component, and the output data variable of the language question-answering model component is determined as the question-answering service result; the question-answering service result for the business question is displayed on the question-answering interaction page. Repetitive tasks can be repeatedly performed through the question-answering service, reducing labor costs and manual intervention. The question-answering service is convenient and easy to use, which can reduce implementation error rates, improve production efficiency, and thus enhance the competitiveness of enterprises. Please refer to Figure 7, which is a flow chart of a method for generating a question-and-answer service provided in an embodiment of the present application. The method for generating a question-and-answer service can be executed by a computer device, and the computer device can be the terminal device 200 or the work server 100 as shown in Figure 1. As shown in Figure 7, the method may include: specifically, the question-and-answer service editing application in the computer device can graphically render and present the function code encapsulation component selected by the user through the graphics rendering engine as shown in Figure 5. For example, the user can select an application trigger component (the application trigger component can be the starting component for the entire question-and-answer service workflow operation. When the user enters a business question on an independent page of the question-and-answer service, the function provided by the question-and-answer service can be started through the application trigger component, and the question-and-answer strategy component in the question-and-answer service can be called.Such as the question judgment subcomponent), the question judgment subcomponent, the frequency control subcomponent, the authority judgment subcomponent and the language question and answer model component, as well as the answer return component (displaying the question and answer service results in the question and answer interaction page), the log recording component (recording the log information of the function code encapsulation component in the question and answer service), the message notification component (displaying the message notification in the question and answer interaction page), the process termination component (the question and answer service has an error or exception, the process is terminated) and the process end component (the question and answer service is completed, the process is ended) and other function code encapsulation components. The processing flow relationship for creating the above-mentioned function code encapsulation components is shown in Figure 7, and the generation of the question and answer service is confirmed by the processing flow relationship of the selected function code encapsulation component. The specific functional implementation method of the application trigger component in the question and answer service can refer to the description of step S202 in the corresponding embodiment of Figure 6, which will not be repeated here. The application trigger component can transmit business problems to the problem judgment subcomponent. The specific functional implementation method of the problem judgment subcomponent, the frequency control subcomponent and the authority judgment subcomponent in the question and answer service can refer to the description of step S204 in the corresponding embodiment of Figure 6, which will not be repeated here. The computer device may determine a "no" process logic for business questions containing prohibited business terms or whose file format is an exception to question handling as "no," terminating the process with a "no" response. Business questions that do not contain prohibited business terms and whose file format is not an exception to question handling may be transmitted to the frequency control component. The computer device may determine a "no" process logic for business questions whose question frequency (frequency of invoking the Q&A service) exceeds the filtering threshold indicated by the frequency control subcomponent, terminating the process with a "no" response. Business questions whose question frequency is less than or equal to the filtering threshold may be transmitted to the permission determination subcomponent. The computer device may determine a "no" process logic for business questions whose question permission does not fall within the question permission scope, display a "no permission to ask" message on the Q&A interaction page, and terminate the process with a "no" response. Business questions that fall within the question permission scope may be transmitted to the language Q&A model component, which performs semantic recognition on the business question to obtain a Q&A service result. The computer device may display the Q&A service result for the business question on the Q&A interaction page via the answer return component and log information from the function code encapsulation component of the Q&A service via the logging component, ultimately terminating the Q&A service process. The embodiments of the present application can provide various types of function code packaging components that support independent deployment, realize flexible access of function code packaging components, and generate question and answer services through the processing flow connection relationship between the function code packaging components selected by the user. It allows people who do not have professional programming knowledge to develop question and answer services through visual function code packaging components and processing flow connection relationships, rather than writing code to develop question and answer services, thereby improving the production efficiency and versatility of question and answer service development, and reducing development costs and implementation error rates.Please refer to Figure 8, which is a schematic diagram of the structure of a generation device for a question-and-answer service provided by an embodiment of the present application. As shown in Figure 8, the generation device 1 for the question-and-answer service includes a training text acquisition module 2001, a model training module 2002, a component display module 2003, a component editing module 2004, a component confirmation module 2005, a first arrangement module 2006, a second arrangement module 2007, a third arrangement module 2008, a fourth arrangement module 2009, a first configuration module 2010, a second configuration module 2011, a third configuration module 2012, a fourth configuration module 2014, a configuration update module 2015, a log table creation module 2016, a log processing module 2017, a first input module 2018, a transmission processing module 2019, a second input module 2020, a question-and-answer processing module 2021, and a result display module 2032. The question-answer processing module includes an execution table creation unit 2022, a first component processing unit 2023, a second component processing unit 2024, a third component processing unit 2025, a statistical information acquisition unit 2026, a first filtering unit 2027, a second filtering unit 2028, a third filtering unit 2029, a preprocessing unit 2030, and a semantic recognition unit 2031. The training text acquisition module 2001 is used to obtain a training text set, filter business stop words in the training text set, and perform text segmentation on the filtered training text set to obtain M text segments, where M is a positive integer; the number of occurrences of each text segment in the M text segments is counted to obtain the text frequencies corresponding to the M text segments. The model training module 2002 is used to obtain an initial language model and establish a text dictionary for the initial language model based on the M text segments and the M text frequencies. The text dictionary includes word vectors corresponding to the M text segments. Obtain question and answer samples, and construct question and answer word vectors corresponding to the question and answer samples through a text dictionary; train the initial language model through the question and answer word vector to obtain a language question and answer model component. The specific functional implementation of the training text acquisition module 2001 and the model training module 2002 can be found in the description of step S101 in the embodiment corresponding to Figure 4, which will not be repeated here. The component display module 2003 is used to display a function code encapsulation component set in the editing interaction page of the question and answer service editing application. The function code encapsulation component set includes multiple function code encapsulation components, and the multiple function code encapsulation components include at least one trigger component, at least one question and answer strategy component, and at least one language question and answer model component. The specific functional implementation of the component display module 2003 can be found in the description of step S101 in the embodiment corresponding to Figure 4, which will not be repeated here.The component editing module 2004 is configured to, in response to a component selection operation for a function code encapsulated component set, display the trigger component, question-and-answer strategy component, and language question-and-answer model component selected by the component selection operation on the question-and-answer service editing page included in the edit interaction page. The question-and-answer service editing page is further configured to display the processing flow connection relationship between the selected trigger component, question-and-answer strategy component, and language question-and-answer model component. The specific functional implementation of the component editing module 2004 can be found in the description of step S102 in the corresponding embodiment of FIG4 and will not be repeated here. The component confirmation module 2005 is configured to, in response to a question-and-answer service confirmation operation triggered on the edit interaction page of the question-and-answer service editing application, generate a question-and-answer service based on the processing flow connection relationship between the selected trigger component, question-and-answer strategy component, and language question-and-answer model component. The trigger component in the question-and-answer service is configured to call the question-and-answer strategy component in the question-and-answer service. The question-and-answer strategy component in the question-and-answer service is configured to filter the business questions after obtaining the business questions. The model component in the question-and-answer service is configured to output the question-and-answer service results based on the filtered business questions. The specific functional implementation of the component confirmation module 2005 can be found in the description of step S103 in the embodiment corresponding to FIG4 , and will not be repeated here. The first orchestration module 2006 is used to respond to the first component orchestration operation on the function code encapsulation component selected in the Q&A service editing page, create a processing flow connection relationship between the selected trigger component and the Q&A strategy component; based on the processing flow connection relationship between the selected trigger component and the Q&A strategy component, determine the selected trigger component as the function code encapsulation component for executing the Q&A service start process, and determine the selected Q&A strategy component as the function code encapsulation component for executing the next process of the trigger component. The second orchestration module 2007 is used to respond to the second component orchestration operation on the function code encapsulation component selected in the Q&A service editing page, create a processing flow connection relationship between the selected Q&A strategy component and the language Q&A model component; based on the processing flow connection relationship between the selected Q&A strategy component and the language Q&A model component, determine the selected language Q&A model component as the function code encapsulation component for executing the next process of the Q&A strategy component. The third orchestration module 2008 is used to respond to the third component orchestration operation for the question judgment subcomponent and the frequency control subcomponent selected in the question and answer service editing page, and create a processing flow connection relationship between the selected question judgment subcomponent and the frequency control subcomponent; based on the processing flow connection relationship between the selected question judgment subcomponent and the frequency control subcomponent, determine the selected question judgment subcomponent as a functional code encapsulation component for executing the question and answer strategy starting process, and determine the selected frequency control subcomponent as a functional code encapsulation component for executing the next process of the question judgment subcomponent.The fourth orchestration module 2009 is used to create a processing flow connection relationship between the selected frequency control subcomponent and the authority judgment subcomponent in response to the fourth component orchestration operation for the frequency control subcomponent and the authority judgment subcomponent selected in the question-and-answer service editing page; based on the processing flow connection relationship between the selected frequency control subcomponent and the authority judgment subcomponent, the selected authority judgment subcomponent is determined as the functional code encapsulation component for executing the next process of the frequency control subcomponent. The specific functional implementation of the first orchestration module 2006, the second orchestration module 2007, the third orchestration module 2008 and the fourth orchestration module 2009 can be found in the description of step S102 in the corresponding embodiment of Figure 4, and will not be repeated here. The first configuration module 2010 is used to configure the business banned words and question handling exception format of the selected question judgment subcomponent in the parameter configuration page in the editing interaction page in response to the first component configuration operation for the question judgment subcomponent selected in the question-and-answer service editing page. The second configuration module 2011 is used to configure the frequency threshold of the selected frequency control subcomponent in the parameter configuration page in response to the second component configuration operation for the frequency control subcomponent selected in the question and answer service editing page. The third configuration module 2012 is used to configure the question permission range of the selected permission judgment subcomponent in response to the third component configuration operation for the permission judgment subcomponent selected in the question and answer service editing page. The fourth configuration module 2014 is used to display multiple instance template identifiers in the parameter configuration page of the editing interaction page in response to the component configuration operation for the function code encapsulation component selected in the question and answer service editing page; and determine the target instance template identifier among the multiple instance template identifiers. The configuration update module 2015 is used to update the selected function code encapsulation component based on the target instance template identifier in the question and answer service editing page. The instance template identifier is used to determine the function function corresponding to the function code encapsulation component, and the function function is used to perform functional processing on business problems. The specific functional implementation of the first configuration module 2010, the second configuration module 2011, the third configuration module 2012, the fourth configuration module 2014 and the configuration update module 2015 can be found in the description of step S102 in the corresponding embodiment of Figure 4, which will not be repeated here. The log table creation module 2016 is used to create an execution log table for the question and answer service based on the selected trigger component, question and answer strategy component and language question and answer model component; and record the log information of the function code encapsulation component in the question and answer service through the execution log table. The log processing module 2017 is used to count the operating status and service quality quantitative data of the question and answer service based on the log information; based on the operating status and service quality quantitative data of the question and answer service, adjust at least one of the trigger component, question and answer strategy component, language question and answer model component or processing flow connection relationship in the question and answer service.The specific functional implementation of the log table creation module 216 and the log processing module 217 can be found in the description of step S103 in the embodiment corresponding to FIG4 and will not be repeated here. The first input module 218 is used to display the question-and-answer interaction page of the question-and-answer service in the communication application and obtain the business question entered in the question-and-answer interaction page, the business question including the trigger parameter; the transmission processing module 219 is used to obtain the message certificate through the trigger component if the trigger parameter meets the question-and-answer trigger condition indicated by the trigger component, and send the business question and the message certificate to the communication server, so that the communication server encrypts the business question according to the message certificate to obtain the encrypted ciphertext, and sends the encrypted ciphertext to the working server. The message certificate is also stored on the working server. The working server is used to decrypt the encrypted ciphertext using the stored message certificate to obtain the business question. The working server is also used to output the question-and-answer service result corresponding to the business question through the question-and-answer service. The specific functional implementation of the first input module 218 and the transmission processing module 219 can be found in the description of step S202 in the embodiment corresponding to FIG6 and will not be repeated here. The second input module 2020 is used to display the Q&A interaction page of the Q&A service in the communication application and obtain the business question entered in the Q&A interaction page, which includes trigger parameters. The specific functional implementation of the second input module 2020 can be found in the description of step S202 in the corresponding embodiment of Figure 6 and will not be repeated here. The Q&A processing module 2021 is used to call the Q&A strategy component in the Q&A service through the trigger component if the trigger parameters meet the Q&A trigger conditions indicated by the trigger component. The Q&A processing module 2021 is also used to filter business questions using the Q&A strategy component in the Q&A service to obtain filtered business questions. In one embodiment, the Q&A processing module 2021 is also used to perform question and answer recognition on the filtered business questions using the language question and answer model component in the Q&A service and output the Q&A service results. In one embodiment, the result display module 2032 is also used to display the Q&A service results for the business questions on the Q&A interaction page. The execution table creation unit 2022 is used to obtain an execution record table for the Q&A service, which includes an instance template identifier. The execution table creation unit 2022 can specifically create an execution record table for the question-and-answer service based on the selected trigger component, question-and-answer strategy component, and language question-and-answer model component, and record the component information of the functional code encapsulation component in the question-and-answer service through the execution record table; the component information includes the processing flow connection relationship, the instance name and instance template identifier used to characterize the functional code encapsulation component in the question-and-answer service; the specific functional implementation method of the execution table creation unit 2022 can be found in the description of step S201 in the corresponding embodiment of Figure 6, which will not be repeated here. The first component processing unit 2023 is used to determine the filtered business question as the output data variable of the trigger component, and determine the output data variable of the trigger component as the input data variable of the question-and-answer strategy component.The first component processing unit 2023 can be specifically used to determine the business problem as the output data variable of the trigger component, and determine the output data variable of the trigger component as the input data variable of the question and answer strategy component according to the processing flow connection relationship indicated by the execution record table; the specific functional implementation of the first component processing unit 2023 can refer to the description of step S203 in the corresponding embodiment of Figure 6, which will not be repeated here. The second component processing unit 2024 is used to determine the functional function corresponding to the question and answer strategy component based on the instance template identifier, and perform functional processing on the input data variable of the question and answer strategy component through the functional function corresponding to the question and answer strategy component to obtain the output data variable of the question and answer strategy component; determine the output data variable of the question and answer strategy component as the input data variable of the language question and answer model component; the specific functional implementation of the second component processing unit 2024 can refer to the description of step S204 in the corresponding embodiment of Figure 6, which will not be repeated here. The third component processing unit 2025 is configured to determine the function corresponding to the language question and answer model component based on the instance template identifier; perform function processing on the input data variables of the language question and answer model component using the function corresponding to the language question and answer model component to obtain the output data variables of the language question and answer model component; and determine the output data variables of the language question and answer model component as the question and answer service result. The third component processing unit 2025 can be specifically configured to determine the function corresponding to the language question and answer model component based on the instance template identifier in the execution record table; perform function processing on the input data variables of the language question and answer model component using the function corresponding to the language question and answer model component to obtain the output data variables of the language question and answer model component; and determine the output data variables of the language question and answer model component as the question and answer service result. The execution record table is also configured to store the input data variables and output data variables of the function code encapsulation component in the question and answer service. The specific functional implementation of the third component processing unit 2025 can be found in the description of step S205 in the embodiment corresponding to FIG6 and will not be repeated here. The statistical information acquisition unit 2026 is used to obtain the question-and-answer statistical information of the user who initiated the business question. The question-and-answer statistical information includes the question frequency and the question authority. The question frequency is the frequency of the user calling the question-and-answer service, and the question authority is determined based on the user's user authority. The first filtering unit 2027 is used to obtain business questions, filter the business questions through the business banned words and question handling exception format indicated by the question judgment subcomponent, and transmit the business questions filtered by the question judgment subcomponent to the frequency control subcomponent. The second filtering unit 2028 is used to filter the business questions filtered by the question judgment subcomponent through the frequency threshold and question frequency indicated by the frequency control subcomponent, and transmit the business questions filtered by the frequency control subcomponent to the authority judgment subcomponent.The third filtering unit 2029 is used to filter the business questions filtered by the frequency control subcomponent based on the question authority range and questioning authority indicated by the authority judgment subcomponent, and obtain filtered business questions. The specific functional implementation of the first filtering unit 2027, the second filtering unit 2028, and the third filtering unit 2029 can be found in the description of step S204 in the embodiment corresponding to Figure 6, and will not be repeated here. The preprocessing unit 2030 is used to remove business stop words in the filtered business questions to obtain updated business questions; perform text segmentation on the updated business questions to obtain N text segments, where N is a positive integer. The semantic recognition unit 2031 is used to convert the N text segments into text word vectors, encode the text word vectors, and obtain contextual information and semantic information of the text word vectors; perform semantic recognition on the contextual information and semantic information through the language question answering model component to obtain a question and answer service result for the business question. The specific functional implementation of the preprocessing unit 2030 and the semantic recognition unit 2031 can be found in the description of step S205 in the embodiment corresponding to Figure 6, and will not be repeated here. The result display module 2032 is used to display the question-and-answer service results for business questions on the question-and-answer interaction page. The specific functional implementation of the result display module 2032 can be found in the description of step S206 in the embodiment corresponding to Figure 6, and will not be repeated here. Please refer to Figure 9, which is a structural diagram of a computer device provided in an embodiment of the present application. As shown in Figure 9, the computer device 1000 may include: a processor 1001, a network interface 1004 and a memory 1005. In addition, the above-mentioned computer device 1000 may also include: a user interface 1003, and at least one communication bus 1002. Among them, the communication bus 1002 is used to realize the connection and communication between these components. Among them, the user interface 1003 may include a display screen (Display), a keyboard (Keyboard), and optionally the user interface 1003 may also include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1005 may be a high-speed RAM memory or a non-volatile memory (non-volatile memory), such as at least one disk storage. The memory 1005 may optionally be at least one storage device located away from the processor 1001. As shown in FIG9 , the memory 1005 as a computer-readable storage medium may include an operating system, a network communication module, a user interface module, and a device control application.In the computer device 1000 shown in FIG9 , the network interface 1004 can provide a network communication network element; the user interface 1003 is mainly used to provide an input interface for the user; and the processor 1001 can be used to call the device control application stored in the memory 1005 to achieve: in the editing interaction page of the question-answering service editing application, a function code encapsulation component set is displayed, the function code encapsulation component set includes a variety of function code encapsulation components, and the variety of function code encapsulation components include at least one trigger component, at least one question-answering strategy component and at least one language question-answering model component; in response to the component selection operation for the function code encapsulation component set, in the question-answering service editing page included in the editing interaction page, the function code encapsulation component set is displayed through the component selection operation. The trigger component, question-and-answer strategy component, and language question-and-answer model component selected by the selection operation are selected. The question-and-answer service editing page is also used to display the processing flow connection relationship between the selected trigger component, question-and-answer strategy component, and language question-and-answer model component; and in response to the question-and-answer service confirmation operation triggered in the editing interaction page of the question-and-answer service editing application, based on the processing flow connection relationship between the selected trigger component, question-and-answer strategy component, and language question-and-answer model component, a question-and-answer service is generated. The trigger component in the question-and-answer service is used to call the question-and-answer strategy component in the question-and-answer service. The question-and-answer strategy component in the question-and-answer service is used to filter the business questions after obtaining the business questions. The model component in the question-and-answer service is used to output the question-and-answer service results based on the filtered business questions. It should be understood that the computer device 1000 described in the embodiment of the present application can execute the description of the generation method of the question-and-answer service in any of the embodiments corresponding to Figures 4, 6, and 7 above, which will not be repeated here. In addition, the description of the beneficial effects of using the same method will not be repeated. In addition, it should be pointed out here that: the embodiment of the present application also provides a computer-readable storage medium, and the computer-readable storage medium stores computer-readable instructions. When the processor executes the computer-readable instructions, it can execute the description of the method for generating the above-mentioned question-and-answer service in the embodiment corresponding to any one of Figures 4, 6 and 7 above. Therefore, it will not be repeated here. In addition, the description of the beneficial effects of adopting the same method will not be repeated. For technical details not disclosed in the embodiment of the computer-readable storage medium involved in this application, please refer to the description of the embodiment of the method of this application. The computer-readable storage medium can be a device for generating the question-and-answer service provided by any of the aforementioned embodiments or an internal storage unit of the computer device, such as a hard disk or memory of a computer device. The computer-readable storage medium can also be an external storage device of the computer device, such as a plug-in hard disk, a smart memory card (smart media card, SMC), a secure digital (secure digital, SD) card, a flash card (flash card), etc. equipped on the computer device.Furthermore, the computer-readable storage medium may also include both an internal storage unit of the computer device and an external storage device. The computer-readable storage medium is used to store the computer-readable instructions and other programs and data required by the computer device. The computer-readable storage medium can also be used to temporarily store data that has been displayed or is to be displayed. In addition, it should be noted here that: the embodiment of the present application also provides a computer program product, which includes computer-readable instructions, and the computer-readable instructions are stored in a computer-readable storage medium. The processor of the computer device reads the computer-readable instructions from the computer-readable storage medium, and the processor executes the computer-readable instructions, so that the computer device executes the method provided in any one of the embodiments corresponding to Figures 4, 6 and 7 above. The terms "first", "second", etc. in the description, claims and drawings of the embodiments of the present application are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, apparatus, product, or device comprising a series of steps or units is not limited to the listed steps or modules, but may optionally include steps or modules not listed, or may optionally include other steps and units inherent to the process, method, apparatus, product, or device. Those skilled in the art will appreciate that the units and algorithmic steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented using electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the above description generally describes the components and steps of each example in terms of network elements. Whether these network elements are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described network elements for each specific application, but such implementations should not be considered beyond the scope of this application. The technical features of the above embodiments may be combined in any manner. To simplify the description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there are no contradictions in the combination of these technical features, they should be considered to be within the scope of this specification. The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art could make various modifications and improvements without departing from the spirit of the present application, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present patent application shall be determined by the appended claims.

Claims

1. A method for generating a question-answering service, executed by a computer device, comprising: In an editing interaction page of a question-answering service editing application, a function code encapsulation component set is displayed, wherein the function code encapsulation component set includes multiple function code encapsulation components, and the multiple function code encapsulation components include at least one trigger component, at least one question-answering strategy component, and at least one language question-answering model component; In response to a component selection operation on the function code encapsulation component set, the trigger component, the question and answer strategy component, and the language question and answer model component selected by the component selection operation are displayed in a question and answer service editing page included in the editing interaction page, and the question and answer service editing page is further used to display a processing flow connection relationship between the selected trigger component, the question and answer strategy component, and the language question and answer model component; and In response to a question and answer service confirmation operation triggered in the editing interaction page of the question and answer service editing application, a question and answer service is generated based on the processing flow connection relationship between the selected trigger component, the question and answer strategy component and the language question and answer model component. The trigger component in the question and answer service is used to call the question and answer strategy component in the question and answer service. The question and answer strategy component in the question and answer service is used to perform question and answer filtering on the business problem after obtaining the business problem. The model component in the question and answer service is used to output the question and answer service result according to the filtered business problem.

2. The method according to claim 1, further comprising: Displaying a question-and-answer interaction page of the question-and-answer service in a communication application, and acquiring a business question input in the question-and-answer interaction page, wherein the business question includes a trigger parameter; If the trigger parameter meets the question and answer trigger condition indicated by the trigger component, the message credential is obtained through the trigger component, and the business question and the message credential are sent to the communication server, so that the communication server encrypts the business question according to the message credential to obtain the encrypted ciphertext, and sends the encrypted ciphertext to the working server. The message credential is also stored in the working server. The working server is used to decrypt the encrypted ciphertext through the stored message credential to obtain the business question. The working server is also used to output the question and answer service result corresponding to the business question through the question and answer service.

3. The method according to claim 1, further comprising: Displaying a question-and-answer interaction page of the question-and-answer service in a communication application, and acquiring a business question input in the question-and-answer interaction page, wherein the business question includes a trigger parameter; If the trigger parameter satisfies the question-and-answer trigger condition indicated by the trigger component, the question-and-answer strategy component in the question-and-answer service is called through the trigger component; Filtering the business questions through the question-answering strategy component in the question-answering service to obtain filtered business questions; Through the language question-answering model component in the question-answering service, question-answering recognition is performed on the filtered business questions, and the question-answering service result is output; The question-and-answer service result for the business question is displayed in the question-and-answer interaction page.

4. The method according to claim 3, further comprising: Obtaining an execution record table of the question-and-answer service, wherein the execution record table includes an instance template identifier; The step of performing question and answer recognition on the filtered business questions through the language question and answer model component in the question and answer service and outputting the question and answer service result includes: Determine the filtered business question as the output data variable of the trigger component, and determine the output data variable of the trigger component as the input data variable of the question-answering strategy component; Determine the function corresponding to the question-and-answer strategy component based on the instance template identifier, and perform function processing on the input data variable of the question-and-answer strategy component through the function corresponding to the question-and-answer strategy component to obtain the output data variable of the question-and-answer strategy component; Determine the output data variable of the question-answering strategy component as the input data variable of the language question-answering model component; Based on the instance template identifier, determining a function corresponding to the language question answering model component; Performing function processing on the input data variables of the language question and answer model component through the function function corresponding to the language question and answer model component to obtain the output data variables of the language question and answer model component; The output data variable of the language question answering model component is determined as the question answering service result.

5. The method according to claim 3, wherein the step of performing question and answer recognition on the filtered business questions through the language question and answer model component in the question and answer service and outputting the question and answer service result comprises: Removing business stop words from the filtered business questions to obtain updated business questions; Performing text segmentation on the update business problem to obtain N text segments, where N is a positive integer; Convert the N text fragments into text word vectors respectively, encode the text word vectors, and obtain context information and semantic information of the text word vectors; The language question-answering model component is used to perform semantic recognition on the context information and the semantic information to obtain a question-answering service result for the business question.

6. According to the method of any one of claims 3 to 5, the question-answering strategy component in the question-answering service includes a question judgment subcomponent, a frequency control subcomponent and an authority judgment subcomponent, and filtering the business questions through the question-answering strategy component in the question-answering service to obtain the filtered business questions includes: Obtaining question-and-answer statistical information of the user who initiated the business question, wherein the question-and-answer statistical information includes question-asking frequency and question-asking authority, wherein the question-asking frequency is the frequency with which the user calls the question-asking service, and the question-asking authority is determined based on the user authority of the user; Acquire the business problem, filter the business problem by using the business banned words and problem handling exception format indicated by the problem judgment subcomponent, and transmit the business problem filtered by the problem judgment subcomponent to the frequency control subcomponent; Filtering the business questions filtered by the question determination subcomponent through the frequency threshold indicated by the frequency control subcomponent and the question frequency, and transmitting the business questions filtered by the frequency control subcomponent to the authority determination subcomponent; The business problems filtered by the frequency control subcomponent are filtered through the question authority range indicated by the authority judgment subcomponent and the questioning authority to obtain filtered business problems.

7. The method according to any one of claims 1 to 6, further comprising: In response to a first component configuration operation for the question determination subcomponent selected in the question-answering service editing page, configuring the business banned words and question handling exception format of the selected question determination subcomponent in a parameter configuration page in the editing interaction page; In response to a second component configuration operation for the frequency control subcomponent selected in the question-and-answer service editing page, configuring a frequency threshold of the selected frequency control subcomponent in the parameter configuration page; In response to the third component configuration operation for the authority determination subcomponent selected in the question-and-answer service editing page, the question authority range of the selected authority determination subcomponent is configured in the parameter configuration page.

8. The method according to any one of claims 1 to 7, further comprising: In response to a component configuration operation for a function code package component selected in the question-answering service editing page, a plurality of instance template identifiers are displayed in a parameter configuration page in the editing interaction page; Determining a target instance template identifier among the multiple instance template identifiers; In the question-and-answer service editing page, the selected function code encapsulation component is updated based on the target instance template identifier, the instance template identifier is used to determine the function function corresponding to the function code encapsulation component, and the function function is used to perform functional processing on the business problem.

9. The method according to any one of claims 1 to 8, further comprising: In response to a first component orchestration operation on a function code package component selected in the question-and-answer service editing page, a processing flow connection relationship is created between the selected trigger component and the question-and-answer strategy component; Based on the processing flow connection relationship between the selected trigger component and the question-and-answer strategy component, the selected trigger component is determined as a function code encapsulation component for executing the question-and-answer service start flow, and the selected question-and-answer strategy component is determined as a function code encapsulation component for executing the next flow of the trigger component; In response to a second component orchestration operation on a function code package component selected in the question-and-answer service editing page, a processing flow connection relationship is created between the selected question-and-answer strategy component and the language question-and-answer model component; Based on the processing flow connection relationship between the selected question and answer strategy component and the language question and answer model component, the selected language question and answer model component is determined as a functional code encapsulation component for executing the next flow of the question and answer strategy component.

10. The method according to any one of claims 1 to 9, wherein the question-answering strategy component comprises a question determination subcomponent, a frequency control subcomponent and an authority determination subcomponent, and the method further comprises: In response to a third component arrangement operation for the question determination subcomponent and the frequency control subcomponent selected in the question-answering service editing page, creating a processing flow connection relationship between the selected question determination subcomponent and the frequency control subcomponent; Based on the processing flow connection relationship between the selected question judgment subcomponent and the frequency control subcomponent, the selected question judgment subcomponent is determined as a functional code encapsulation component for executing the question-answering strategy start flow, and the selected frequency control subcomponent is determined as a functional code encapsulation component for executing the next flow of the question judgment subcomponent; In response to the fourth component arrangement operation for the frequency control subcomponent and the authority judgment subcomponent selected in the question-and-answer service editing page, a processing flow connection relationship is created between the selected frequency control subcomponent and the authority judgment subcomponent; Based on the processing flow connection relationship between the selected frequency control subcomponent and the authority judgment subcomponent, the selected authority judgment subcomponent is determined as a functional code encapsulation component for executing the next flow of the frequency control subcomponent.

11. The method according to any one of claims 1 to 10, further comprising: Obtain a training text set, filter business stop words in the training text set, and perform text segmentation on the filtered training text set to obtain M text segments, where M is a positive integer; Counting the number of occurrences of each of the M text segments to obtain text frequencies corresponding to the M text segments; Acquire an initial language model, and establish a text dictionary of the initial language model based on the M text segments and the M text frequencies, wherein the text dictionary includes word vectors corresponding to the M text segments respectively; Obtain a question and answer sample, and construct a question and answer word vector corresponding to the question and answer sample through the text dictionary; The initial language model is trained using the question-answer word vector to obtain the language question-answer model component.

12. The method according to any one of claims 1 to 11, further comprising: Based on the selected trigger component, question-answering strategy component, and language question-answering model component, create an execution log table for the question-answering service; Recording log information of the function code encapsulation component in the question-and-answer service through the execution log table; Collecting quantitative data on the operation status and service quality of the question-and-answer service according to the log information; Based on the running status and the service quality quantification data of the question and answer service, at least one of the trigger component, the question and answer strategy component, the language question and answer model component or the processing flow connection relationship in the question and answer service is adjusted.

13. A device for generating a question-answering service, comprising: A component display module, configured to display a function code encapsulation component set in an editing interaction page of a question-answering service editing application, wherein the function code encapsulation component set includes a plurality of function code encapsulation components, wherein the plurality of function code encapsulation components include at least one trigger component, at least one question-answering strategy component, and at least one language question-answering model component; a component editing module, configured to, in response to a component selection operation on the function code encapsulation component set, display the trigger component, the question and answer strategy component, and the language question and answer model component selected by the component selection operation in a question and answer service editing page included in the editing interaction page, wherein the question and answer service editing page is further configured to display a processing flow connection relationship between the selected trigger component, the question and answer strategy component, and the language question and answer model component; A component confirmation module is used to respond to a question-and-answer service confirmation operation triggered in the editing interaction page of the question-and-answer service editing application, and generate a question-and-answer service based on the processing flow connection relationship between the selected trigger component, the question-and-answer strategy component, and the language question-and-answer model component. The trigger component in the question-and-answer service is used to call the question-and-answer strategy component in the question-and-answer service. The question-and-answer strategy component in the question-and-answer service is used to perform question-and-answer filtering on the business problem after obtaining the business problem. The model component in the question-and-answer service is used to output the question-and-answer service result according to the filtered business problem.

14. A computer device, comprising a processor and a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the processor implements the method according to any one of claims 1 to 12.

15. A computer-readable storage medium storing computer-readable instructions, wherein when the computer-readable instructions are executed by a processor, the processor implements the method according to any one of claims 1 to 12.

16. A computer program product, comprising computer-readable instructions, which, when executed by a processor, enable the processor to implement the method according to any one of claims 1 to 12.

Images