A reply generation method based on forced reading rules and related products

Abstract

The application discloses a reply generation method based on forced reading rules and related products, the method comprises the following steps: obtaining an input request of a user; in the case that the input request contains a preset trigger word, forcibly reading a target rule file under a preset path; based on a priority label in the target rule file, generating an inspection list containing multiple inspection items; embedding the inspection list in a response context corresponding to the input request, the response context is used for generating reply content of the input request, and the reply content carries execution results of the multiple inspection items in the inspection list. In this way, the inspection list is generated based on the priority label in the target rule file, so that the rule is converted into inspection items that can be checked item by item. Furthermore, the inspection list is embedded in the response context, so that the reply content is forced to carry the execution results of each inspection item, so that the user can view the execution of each inspection item at the same time of obtaining the reply content, and effective tracking of the execution process is realized.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a response generation method and related products based on forced reading rules. Background Technology

[0002] In recent years, AI-assisted software development technology has been widely applied, particularly in areas such as interface test case generation and automatic code generation. AI models can automatically generate corresponding output content based on user input requests. Currently, the common approach is to provide the AI ​​model with system prompts or configuration files, incorporating binding rules such as coding standards or testing specifications to guide the AI ​​in generating responses according to preset standards. Based on the received prompts and user input requests, the AI ​​model uses its internal natural language understanding and generation capabilities to output the corresponding response content.

[0003] However, the AI ​​model can only output the final generated response content, not information about the processing of each rule. Users only receive the final response content and cannot confirm whether the AI ​​executed the preset rules or whether any errors or omissions occurred during the execution process. When there are defects in the response content, the lack of records of the execution process makes it difficult for users to trace the root cause of the defects.

[0004] Therefore, how to verify the execution process of AI-generated responses has become an urgent technical problem to be solved. Summary of the Invention

[0005] This application provides a response generation method and related products based on mandatory read rules. A checklist is generated based on priority tags in a target rule file, transforming the rules into verifiable check items. Furthermore, the checklist is embedded in the response context, mandating that the response content carry the execution results of each check item. This allows users to view the execution status of each check item simultaneously with the received response, achieving effective traceability of the execution process.

[0006] In a first aspect, embodiments of this application provide a response generation method based on forced reading rules, the method comprising: Obtain user input requests; If the input request contains a preset trigger word, the target rule file under the preset path will be read forcibly; Based on the priority tags in the target rule file, a checklist containing multiple check items is generated; The checklist is embedded in the response context corresponding to the input request. The response context is used to generate the response content of the input request, and the response content carries the execution results of multiple check items in the checklist.

[0007] One feasible implementation involves generating a checklist containing multiple check items based on priority tags in the target rule file, including: Based on the priority tags in the target rule file, a list of blocking rules is generated; Based on the blocking rule list, a checklist containing multiple check items is generated.

[0008] One feasible implementation, wherein when the input request contains a preset trigger word, forcibly reading a target rule file under a preset path includes: If the input request contains a preset trigger word, the input request will be intercepted based on a preset hook interception mechanism; In response to the interception operation, the target rule file under the preset path is forcibly read.

[0009] One feasible implementation is that the target rule file includes multiple rules, and each of the multiple rules is configured with a priority label; The priority labels include blocking type labels, mandatory type labels, and suggestion type labels.

[0010] One feasible implementation involves generating a checklist containing multiple check items based on priority tags in the target rule file, including: Based on the priority tags in the target rule file, target rules configured with the blocking type tag are selected from the multiple rules; Based on the selected target rules, a checklist containing multiple check items is generated.

[0011] One feasible implementation of the method further includes: If the input request does not contain the preset trigger word, a response is generated based on the input request, and the response does not carry the checklist.

[0012] Secondly, embodiments of this application provide a response generation apparatus based on forced reading rules, comprising: The request retrieval module is used to retrieve user input requests; The forced reading module is used to forcibly read the target rule file under the preset path when the input request contains a preset trigger word; The checklist generation module is used to generate a checklist containing multiple check items based on the priority tags in the target rule file. The response generation module is used to embed the checklist into the response context corresponding to the input request. The response context is used to generate the response content of the input request, and the response content carries the execution results of multiple check items in the checklist.

[0013] One feasible implementation is that the inventory generation module is specifically used for: Based on the priority tags in the target rule file, a list of blocking rules is generated; Based on the blocking rule list, a checklist containing multiple check items is generated.

[0014] Thirdly, embodiments of this application provide an electronic device, the device including: a processor, a memory, and a system bus; The processor and the memory are connected via the system bus; The memory is used to store a program, the program including instructions, which, when executed by the processor, cause the processor to perform any of the implementation steps of the response generation method based on the forced read rule described above.

[0015] Fourthly, embodiments of this application provide a computer-readable storage medium for storing a computer program, which, when executed by a terminal device, implements any of the implementation steps of the above-described response generation method based on forced reading rules.

[0016] As can be seen from the above technical solutions, the embodiments of this application have the following advantages: As can be seen from the above technical solution, this application provides a response generation method and related products based on forced reading rules. The method includes: First, obtaining a user's input request. Next, if the input request contains a preset trigger word, forcibly reading a target rule file under a preset path. Subsequently, based on the priority tags in the target rule file, generating a checklist containing multiple check items. Finally, embedding the checklist into the response context corresponding to the input request, the response context being used to generate the response content of the input request, the response content carrying the execution results of multiple check items in the checklist.

[0017] As can be seen, this application prevents rules from being missed by forcibly reading the target rule file when the user's input request contains preset trigger words. Based on this, a checklist is generated using priority tags in the target rule file, transforming the rules into verifiable check items. Furthermore, the checklist is embedded in the response context, forcibly including the execution results of each check item in the response content. Thus, the user can view the execution status of each check item simultaneously with the received response, achieving effective traceability of the execution process. Attached Figure Description

[0018] Figure 1 A flowchart illustrating a response generation method based on forced reading rules, provided for an embodiment of this application; Figure 2 This is a schematic diagram of a response generation device based on forced reading rules, provided in an embodiment of this application. Detailed Implementation

[0019] As mentioned above, existing technologies have at least the following shortcomings: First, the scattered storage of rules makes it difficult for AI models to obtain all rules at once, easily leading to the omission of key rules; Second, AI models decide whether to follow and how to execute rules based on their own judgment, but existing AI models have limited instruction-following capabilities and cannot reliably distinguish between rules that must be executed and rules that are suggested to be executed, resulting in key rules being ignored or incompletely executed; Third, the priority definitions in the rule files are unclear, with multiple priority marking methods coexisting, making it impossible for AI models to determine the execution order of rules, leading to chaotic execution logic; Fourth, AI models only output the final generated content, and users cannot verify whether the AI ​​model has correctly executed each rule, making the execution process untraceable; Fifth, after an AI model fails to execute due to the omission of rules, the same error will be repeated in the next execution due to the lack of an error feedback mechanism.

[0020] To address the aforementioned issues, this application provides a response generation method and related products based on forced reading rules. The method includes: first, obtaining a user's input request; then, if the input request contains a preset trigger word, forcibly reading a target rule file located at a preset path; subsequently, generating a checklist containing multiple check items based on priority tags in the target rule file; and finally, embedding the checklist into the response context corresponding to the input request. The response context is used to generate a response to the input request, and the response content carries the execution results of multiple check items in the checklist.

[0021] As can be seen, this application avoids rule omissions by forcibly reading the target rule file when the user's input request contains preset trigger words. Based on this, a checklist is generated using priority tags in the target rule file, transforming the rules into verifiable check items. Furthermore, the checklist is embedded in the response context, forcibly including the execution results of each check item in the response content. Thus, users can view the execution status of each check item simultaneously with the received response, achieving effective traceability of the execution process and solving the technical problems of existing technologies where users cannot confirm whether the AI ​​has correctly executed the rules and where the execution process is untraceable.

[0022] Furthermore, to address the issue of missing critical rules due to scattered rule storage, this application forces the reading of target rule files under a preset path, ensuring a unified source and clear path for rules, thus avoiding the omission of critical rules caused by AI models obtaining rules from multiple locations.

[0023] To address the issue of key rules being ignored due to AI models' self-judgment of whether to follow rules, this application triggers a forced reading process by pre-setting trigger words. This ensures that rule acquisition does not depend on the AI ​​model's own judgment, but is enforced by the system, thereby preventing the AI ​​model from mistakenly ignoring key rules due to its limited understanding.

[0024] To address the issue of unclear priority definitions leading to chaotic execution logic, this application generates a checklist based on the priority tags in the target rule file, ensuring that each rule is included in the checklist in an orderly manner according to its priority tag, thus avoiding confusion in the execution order.

[0025] To address the issue of recurring errors, this application mandates that the execution results of each check item be included in the response content. This allows users to promptly identify omissions or errors during the execution process and make corrections, thereby preventing the same errors from recurring in subsequent executions.

[0026] It should be noted that the implementation subject of the response generation method based on forced reading rules in this application is not limited. For example, the response generation method based on forced reading rules in this application can be applied to information processing devices such as servers or terminal devices. The server can be a standalone server, a cluster server, or a cloud server. The terminal device can be an electronic device such as a smartphone, computer, personal digital assistant (PDA), or tablet computer.

[0027] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0028] Figure 1 This is a flowchart illustrating a response generation method based on forced reading rules, provided as an embodiment of this application. (In conjunction with...) Figure 1 As shown, it may include steps S101-S104.

[0029] S101: Obtain user input request.

[0030] In this embodiment of the application, the input request is a natural language instruction or command entered by the user in the AI-assisted software development system. The input request includes user intent information and may include preset trigger words.

[0031] Optionally, the input request can be a piece of text entered by the user in the interactive interface, such as "Please generate test cases for a certain interface" or "Please generate test code based on the following interface definition". The input request contains user intent information and may include preset trigger words.

[0032] In this embodiment, obtaining user input requests can be achieved in various ways, including but not limited to receiving user-submitted request information through a web interface, command-line interface, or application programming interface. In a specific example, after the user enters request text on the front-end interface, the front-end encapsulates the request text into a request message and sends it to the back-end server. The back-end server parses the request message to obtain the user's input request.

[0033] It should be noted that the input requests in this embodiment are presented in a structured or unstructured form. Optionally, the input request can be in plain text format, carrying a description of the user's requirements for the target interface or target function.

[0034] S102: If the input request contains a preset trigger word, force the reading of the target rule file under the preset path.

[0035] In this embodiment, after obtaining the user's input request, it is first determined whether the input request contains a preset trigger word. The preset trigger word is a pre-defined keyword or phrase used to trigger a forced reading process, such as generating test cases or interface testing. By setting the trigger word, this embodiment can accurately identify the user's intent and only activate the process when a rule-based forced process needs to be executed, avoiding unnecessary processing for regular requests.

[0036] Specifically, if the input request contains a preset trigger word, the input request will be intercepted based on a preset hook interception mechanism. The hook interception mechanism is a programmatic interception technique used to insert an interception point in the input request processing flow to trigger a subsequent forced reading process.

[0037] In this embodiment, after intercepting an input request, the hook interception mechanism further determines the scenario type corresponding to the input request. Specifically, when the hook interception mechanism detects that the input request contains a preset trigger word, it determines the current scenario as the first scenario (scenario="A"). At this time, the hook interception mechanism has already output a checklist template, and the AI ​​model does not need any additional operation; it directly performs subsequent processing based on the checklist template. When the input request does not contain a preset trigger word, it determines the current scenario as the second scenario (scenario="B"). At this time, the user has provided complete information, and the AI ​​model calls the corresponding skill for processing. In other words, by distinguishing scenario types, the AI ​​model executes corresponding processing logic in different scenarios, avoiding the AI ​​model from repeatedly operating on the already output checklist template.

[0038] In response to the aforementioned interception operation, the target rule file under a preset path is forcibly read. The preset path is a pre-defined storage path for the target rule file, such as " / config / MANDATORY-RULES.md". The preset path can be an absolute path or a relative path; this application does not limit this. In this embodiment, the target rule file is a pre-configured text file containing multiple rules, and each rule has a priority label. After forcibly reading the target rule file, the hook interception mechanism also outputs a P0-level checklist table. This P0-level checklist table displays the checks to be executed to the user, allowing the user to know the contents of the checks to be executed before the AI ​​model executes.

[0039] Therefore, by bypassing the AI ​​model's own judgment logic and proactively reading the target rule file from a preset path, the system no longer relies on the AI ​​model to decide whether to read it. This forced reading operation ensures that multiple rules in the target rule file are reliably incorporated into the subsequent processing flow, avoiding rule omissions due to AI model errors. Simultaneously, by outputting a P0-level checklist, users can preview the upcoming checks before execution, improving the transparency of the execution process.

[0040] Furthermore, in the forced reading process of this application embodiment, it is not allowed to read other files to replace the target rule file, nor is it allowed to simplify or truncate the content of the target rule file, so that the rules on which subsequent processing is based are complete, accurate, and without omissions. At the same time, the hook interception mechanism is not allowed to be simplified, that is, the interception operation is not allowed to be skipped or simplified, so that the forced reading process that bypasses the judgment of the AI ​​model can be reliably executed.

[0041] S103: Generate a checklist containing multiple check items based on the priority tags in the target rule file.

[0042] In this embodiment, after reading the target rule file, the file is first parsed to extract multiple rules and their respective priority tags. The priority tags indicate the execution priority of the rule and include three levels: blocking type, mandatory type, and suggestion type. A blocking type tag indicates that the rule is a blocking rule; if it fails to execute correctly, the subsequent process cannot continue. A mandatory type tag indicates that the rule is a mandatory rule; if it fails to execute correctly, the error must be recorded, but the subsequent process can continue. A suggestion type tag indicates that the rule is a suggestion rule, and the AI ​​model can determine whether to execute it.

[0043] Specifically, a blocking rule list is generated based on the priority tags in the target rule file. This blocking rule list is a collection of rules selected from the target rule file that correspond to blocking type tags. In other words, rules configured with blocking type tags are selected from multiple rules, and these selected rules are combined into the blocking rule list. By selecting rules with blocking type tags, critical rules that must be executed and whose failure must halt subsequent processes can be accurately identified.

[0044] Furthermore, based on the blocking rule list, a checklist containing multiple check items is generated. The checklist is a list formed by converting each rule in the blocking rule list into a structured check item.

[0045] Optionally, the checklist is in tabular form, with the table header containing a rule identifier and a check item description to identify each check item to be performed. The rule identifier uniquely identifies the rule number corresponding to the check item; the check item description describes the specific check content of the rule, such as whether the interface type has been identified or whether module name conversion has been completed.

[0046] In this embodiment, a blocking rule list is generated based on priority tags, and then a checklist is generated based on the blocking rule list. This ensures that the highest priority rules are included in the verification scope first, preventing critical rules from being overwhelmed by too many check items. Simultaneously, the rules are transformed into structured, verifiable items, providing a basis for subsequent verification of the execution process.

[0047] S104: Embed the checklist into the response context corresponding to the input request. The response context is used to generate the response content of the input request, and the response content carries the execution results of multiple check items in the checklist.

[0048] In this embodiment, the checklist is embedded in the response context corresponding to the input request. The response context is a data set used to generate the response content, containing various information required to generate the response content. As part of the response context, the checklist is used to instruct the AI ​​model to execute the rules corresponding to each check item in the checklist when generating the response content.

[0049] It should be noted that the checklist in this embodiment is in tabular form, specifically including a rule column, a check item column, an execution result column, and a status column. The rule column identifies the rule number corresponding to the check item, such as "Rule 1" or "Rule 2"; the check item column describes the specific check content of the rule, such as "whether the interface type has been identified" or "whether the module name conversion has been completed"; the execution result column records the specific output content after the check item is executed, such as "interface type: RESTful" or "module name: user-service"; and the status column records the completion status of the check item, such as "executed", "not executed", or "execution failed".

[0050] The response context is used to generate the response content to the input request. Specifically, the response context, carrying a checklist, is input into the AI ​​model, which then generates the response content corresponding to the input request based on the response context. Simultaneously, during the response content generation process, the AI ​​model executes each check item in the checklist and populates the checklist with the execution results of each check item to obtain the execution result. The execution result includes the execution status of each check item, such as executed, not executed, or executed unsuccessfully, as well as the specific output content after each check item is executed.

[0051] It should be noted that the response content generated in this embodiment carries the execution results of multiple check items in the checklist. Optionally, the response content includes the original response content generated by the AI ​​model based on the input request, such as interface test case code, interface test report, etc. Simultaneously, the response content also includes a checklist of execution results. This checklist is output to the user along with the response content, allowing the user to view the execution status of each check item. In other words, the user can simultaneously obtain the execution results of each check item upon receiving the response content, thereby effectively tracing and verifying the rule execution status of the AI ​​model during the response content generation process.

[0052] In this embodiment, the response must include a checklist table filled with execution results. Simply outputting text descriptions instead of a table is not permitted, nor is simplifying the table content or omitting some check items. Furthermore, merely describing the execution status without outputting a structured table is also prohibited. By mandating the output of the checklist table, users can directly view the execution results of each check item, enabling effective traceability of the execution process, timely detection of omissions or errors, and ensuring the quality and verifiability of rule execution.

[0053] In addition, if the input request does not contain a preset trigger word, a response is generated based on the input request, and the response does not include a checklist.

[0054] Based on the steps S101-S104 above, this application provides a response generation method and related products based on forced reading rules. The method includes: First, obtaining a user's input request. Next, if the input request contains a preset trigger word, forcibly reading a target rule file under a preset path. Then, based on priority tags in the target rule file, generating a checklist containing multiple check items. Finally, embedding the checklist into the response context corresponding to the input request. The response context is used to generate the response content of the input request, and the response content carries the execution results of multiple check items in the checklist.

[0055] As can be seen, this application prevents rules from being missed by forcibly reading the target rule file when the user's input request contains preset trigger words. Based on this, a checklist is generated using priority tags in the target rule file, transforming the rules into verifiable check items. Furthermore, the checklist is embedded in the response context, forcibly including the execution results of each check item in the response content. Thus, the user can view the execution status of each check item simultaneously with the received response, achieving effective traceability of the execution process.

[0056] Figure 2 This is a schematic diagram of a response generation device based on forced reading rules, provided as an embodiment of this application. (Combined with...) Figure 2 As shown, the response generation device 200 based on forced reading rules provided in this application embodiment may include: The request retrieval module 201 is used to retrieve user input requests; Forced reading module 202 is used to force read the target rule file under the preset path when the input request contains a preset trigger word; The checklist generation module 203 is used to generate a checklist containing multiple check items based on the priority tags in the target rule file. The response generation module 204 is used to embed the checklist into the response context corresponding to the input request. The response context is used to generate the response content of the input request, and the response content carries the execution results of multiple check items in the checklist.

[0057] Optionally, the inventory generation module 203 is specifically used for: Based on the priority tags in the target rule file, a list of blocking rules is generated; Based on the blocking rule list, a checklist containing multiple check items is generated.

[0058] Optionally, the forced read module 202 is specifically used for: If the input request contains a preset trigger word, the input request will be intercepted based on a preset hook interception mechanism; In response to the interception operation, the target rule file under the preset path is forcibly read.

[0059] Optionally, the target rule file includes multiple rules, and each of the multiple rules is configured with a priority label; The priority labels include blocking type labels, mandatory type labels, and suggestion type labels.

[0060] Optionally, the inventory generation module 203 is specifically used for: Based on the priority tags in the target rule file, target rules configured with the blocking type tag are selected from the multiple rules; Based on the selected target rules, a checklist containing multiple check items is generated.

[0061] Optionally, the response generation device 200 based on forced reading rules may include: If the input request does not contain the preset trigger word, a response is generated based on the input request, and the response does not carry the checklist.

[0062] Furthermore, embodiments of this application also provide an electronic device, including: a processor, a memory, and a system bus; The processor and the memory are connected via the system bus; The memory is used to store one or more programs, the one or more programs including instructions that, when executed by the processor, cause the processor to perform any of the implementation steps of the above-described tamper-proof lottery method.

[0063] Furthermore, embodiments of this application also provide a computer-readable storage medium for storing a computer program, which, when executed by a terminal device, implements any of the implementation steps of the above-described tamper-proof lottery method.

[0064] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that all or part of the steps in the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, a server, or a network communication device such as a media gateway, etc.) to execute the methods described in various embodiments or some parts of the embodiments of this application. It should be noted that the various embodiments in this specification are described in a progressive manner, and each embodiment focuses on describing the differences from other embodiments. The same or similar parts between the various embodiments can be referred to mutually.

[0065] The system disclosed in the embodiments is described in a relatively simple manner because it corresponds to the method disclosed in the embodiments. For relevant details, please refer to the method section.

[0066] It should also be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0067] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A response generation method based on forced reading rules, characterized in that, The method includes: Obtain user input requests; If the input request contains a preset trigger word, the target rule file under the preset path will be read forcibly; Based on the priority tags in the target rule file, a checklist containing multiple check items is generated; The checklist is embedded in the response context corresponding to the input request. The response context is used to generate the response content of the input request, and the response content carries the execution results of multiple check items in the checklist.

2. The method according to claim 1, characterized in that, The step of generating a checklist containing multiple check items based on the priority tags in the target rule file includes: Based on the priority tags in the target rule file, a list of blocking rules is generated; Based on the blocking rule list, a checklist containing multiple check items is generated.

3. The method according to claim 1, characterized in that, When the input request contains a preset trigger word, the forced reading of the target rule file under the preset path includes: If the input request contains a preset trigger word, the input request will be intercepted based on a preset hook interception mechanism; In response to the interception operation, the target rule file under the preset path is forcibly read.

4. The method according to claim 1, characterized in that, The target rule file includes multiple rules, and each rule is configured with a priority label; The priority labels include blocking type labels, mandatory type labels, and suggestion type labels.

5. The method according to claim 4, characterized in that, The step of generating a checklist containing multiple check items based on the priority tags in the target rule file includes: Based on the priority tags in the target rule file, target rules configured with the blocking type tag are selected from the multiple rules; Based on the selected target rules, a checklist containing multiple check items is generated.

6. The method according to claim 1, characterized in that, The method further includes: If the input request does not contain the preset trigger word, a response is generated based on the input request, and the response does not carry the checklist.

7. A response generation device based on forced reading rules, characterized in that, include: The request retrieval module is used to retrieve user input requests; The forced reading module is used to forcibly read the target rule file under the preset path when the input request contains a preset trigger word; The checklist generation module is used to generate a checklist containing multiple check items based on the priority tags in the target rule file. The response generation module is used to embed the checklist into the response context corresponding to the input request. The response context is used to generate the response content of the input request, and the response content carries the execution results of multiple check items in the checklist.

8. The apparatus according to claim 7, characterized in that, The inventory generation module is specifically used for: Based on the priority tags in the target rule file, a list of blocking rules is generated; Based on the blocking rule list, a checklist containing multiple check items is generated.

9. An electronic device, characterized in that, The device includes: a processor, a memory, and a system bus; The processor and the memory are connected via the system bus; The memory is used to store a program, the program including instructions that, when executed by the processor, cause the processor to perform the steps of any one of claims 1-6.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store a computer program, which, when executed by a terminal device, implements the steps of the response generation method based on forced reading rules as described in any one of claims 1-6.

Images