Dialogue control method and system based on artificial intelligence dialogue large model

Abstract

The invention discloses a dialogue control method and system based on an artificial intelligence dialogue large model, and the method comprises the following steps: S10, providing a dialogue background description for the dialogue large model, and enabling the dialogue background description to be used for setting a dialogue scene; s20, executing a flow of a full-field open-field white stage; s30, executing a process of a paragraph opening white stage; s40, executing the process of ending the open-field white stage; and S50, proposing a large model summary requirement, and generating a dialogue summary according to the large model summary requirement. The embodiment of the invention provides a dialogue process control method based on an artificial intelligence dialogue large model. The process can actively guide and manage multi-round and multi-theme dialogues with users. According to the system, an artificial intelligence dialogue large model can be converted from a passive information provider to an active dialogue guider, and the system has important value in occasions needing deep communication such as psychological counseling.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence dialogue, and in particular to a dialogue control method and system based on an artificial intelligence dialogue large model. Background Art

[0002] In the current AI dialogue big model, users usually take the initiative to ask questions or make demands, and the AI ​​dialogue big model generates corresponding answers after receiving the questions.

[0003] Although this mode is suitable for general information retrieval or daily communication, in certain specific vertical fields, such as psychological consultation and medical consultation, this passive dialogue mode may not be efficient or in-depth enough. Especially in cases that require professional guidance and in-depth discussion, existing artificial intelligence dialogue systems usually seem to be unable to cope with it. At the same time, in these specific vertical fields, the purpose of the dialogue is often to obtain sufficient information from the dialogue for subsequent use or judgment, and there are several topics for the dialogue, which makes it difficult to control the artificial intelligence one by one to communicate and collect data. Summary of the invention

[0004] In view of this, an embodiment of the present invention provides a dialogue control method and system based on an artificial intelligence dialogue big model to solve the above technical problems.

[0005] To achieve the above objectives, in a first aspect, a dialogue control method based on an artificial intelligence dialogue model is provided, which comprises the following steps:

[0006] S10: Provides a dialogue background description for the dialogue model to set the dialogue scene;

[0007] S20: executing the process of the full opening speech stage, which includes: providing the full opening speech; obtaining the first user's reply corresponding to the full opening speech; putting the dialogue related to the full opening speech into the full opening speech prompt words of the user's reply; judging whether the first user's reply has met the requirements of the full opening speech prompt words; when it is determined that the first user's reply has met the requirements of the full opening speech prompt words, entering the paragraph opening speech stage;

[0008] S30: Execute the process of the paragraph opening stage, which includes: providing the paragraph opening; obtaining the second user's reply corresponding to the paragraph opening; putting the dialogue related to the paragraph opening into the paragraph prompt words of the user's reply; judging whether the second user's reply has met the requirements of the paragraph prompt words; when it is determined that the second user's reply has met the requirements of the paragraph prompt words, further judging whether there are still necessary paragraphs or whether there are conditional paragraphs that meet the triggering conditions, if the judgment result is no, entering the end of the opening stage;

[0009] S40: executing the process of the end opening speech stage, which includes: providing the end opening speech; obtaining the third user's reply corresponding to the end opening speech; putting the dialogue related to the end opening speech into the end opening speech prompt word of the user's reply; judging whether the third user's reply has met the end opening speech prompt word requirement; when the third user's reply has met the end opening speech prompt word requirement, providing the end speech;

[0010] S50: Propose a large model summary requirement, and generate a dialogue summary according to the large model summary requirement.

[0011] On the other hand, a dialogue control system based on an artificial intelligence dialogue model is provided, the system comprising:

[0012] The dialogue background setting module is used to provide dialogue background description for the dialogue model to set the dialogue scene;

[0013] The full opening speech execution module is used to execute the process of the full opening speech stage, including providing the full opening speech, obtaining the first user's reply corresponding to the full opening speech, putting the dialogue related to the full opening speech into the full opening speech prompt words of the user's reply, judging whether the first user's reply has met the requirements of the full opening speech prompt words, and triggering the paragraph opening speech execution module when it is determined that the first user's reply has met the requirements of the full opening speech prompt words;

[0014] A paragraph opening statement execution module is used to execute the process of the paragraph opening statement stage, including providing a paragraph opening statement, obtaining a second user reply corresponding to the paragraph opening statement, putting the paragraph opening statement related dialogue into the paragraph prompt word of the user reply, judging whether the second user reply has met the paragraph prompt word requirement, and when it is determined that the second user reply has met the paragraph prompt word requirement, further judging whether there are still necessary paragraphs or whether there are conditional paragraphs that meet the triggering condition, and if the judgment result is no, triggering the end of the opening statement execution module;

[0015] an end remark execution module, used to execute the process of the end remark stage, including providing the end remark, obtaining the third user's reply corresponding to the end remark, putting the dialogue related to the end remark into the end remark prompt word of the user's reply, judging whether the third user's reply has met the end remark prompt word requirement, and providing the end remark when the third user's reply has met the end remark prompt word requirement;

[0016] The summary generation module is used to propose a large model summary requirement and generate a dialogue summary according to the large model summary requirement.

[0017] The above technical solution has the following beneficial technical effects:

[0018] The embodiment of the present invention proposes a dialogue process control method based on a large artificial intelligence dialogue model (such as a GPT series model), which can actively guide and manage multi-round, multi-topic dialogues with users.

[0019] This system can transform the artificial intelligence dialogue model from a passive information provider to an active dialogue guide, which is of great value in situations such as psychological counseling that require in-depth communication.

[0020] By intelligently monitoring and analyzing user responses, the system can automatically switch between different preset topics or questions to achieve more targeted conversations.

[0021] The embodiment of the present invention provides a new proactive consultation dialogue process control mechanism with preset chain prompt words based on the artificial intelligence dialogue big model, which is expected to play an important role in psychological consultation, medical consultation and other professional consultation fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The accompanying drawings are used to better understand the present invention and do not constitute an improper limitation of the present invention.

[0023] Figure 1 is a flow chart of a dialogue control method based on an artificial intelligence dialogue model according to an embodiment of the present invention;

[0024] Figure 2 is a flow chart of another dialogue control method based on an artificial intelligence dialogue model according to an embodiment of the present invention;

[0025] Figure 3 It is a functional block diagram of a dialogue control system based on an artificial intelligence dialogue model according to an embodiment of the present invention;

[0026] Figure 4 It is a schematic diagram of the structure of a computer system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0027] The following is a description of exemplary embodiments of the present invention in conjunction with the accompanying drawings, including various details of the embodiments of the present invention to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0028] The embodiment of the present invention proposes a dialogue process control method based on an artificial intelligence dialogue model (such as the GPT series model), which can actively guide and manage multi-round, multi-topic dialogues with users. In short, the system presets a series of dialogues and questions designed by professionals to guide users into dialogues. The system can also automatically adjust the next questions and topics based on the user's response, thereby achieving more targeted and in-depth communication.

[0029] The implementation of this system relies on a carefully designed dialogue framework, which includes various elements related to dialogue management, such as dialogue requirements, turns, opening remarks, preset questions, judgment conditions, etc., to support flexible and automated dialogue processes, and put prompts designed by different fields or professionals in this framework. As long as these prompts meet the requirements of the dialogue process control mechanism, the required data and information can be collected one by one for use in subsequent processes.

[0030] The purpose of the preset chain prompt active consultation dialogue process control mechanism based on the artificial intelligence dialogue big model is to use the preset framework content to guide the dialogue of the generative big model and collect sufficient information for subsequent business or other practical scenarios.

[0031] Figure 1 is a flow chart of a dialogue control method based on an artificial intelligence dialogue model according to an embodiment of the present invention, such as Figure 1 As shown, the method comprises the following steps:

[0032] S10: Provides a dialogue background description for the dialogue model to set the dialogue scene;

[0033] S20: executing the process of the full opening speech stage, which includes: providing the full opening speech; obtaining the first user's reply corresponding to the full opening speech; putting the dialogue related to the full opening speech into the full opening speech prompt words of the user's reply; judging whether the first user's reply has met the requirements of the full opening speech prompt words; when it is determined that the first user's reply has met the requirements of the full opening speech prompt words, entering the paragraph opening speech stage;

[0034] Specifically, the conversation related to the full opening remarks includes the conversation content between the full opening remarks and the user's reply to the full opening remarks.

[0035] Specifically, putting the conversation related to the opening remarks into the opening remarks prompt of the user's reply means that in the dialogue system or dialogue model, after the user replies to the opening remarks, the system will integrate the conversation content related to the opening remarks and place it in the user's reply as a prompt (or guide). This helps to maintain the continuity of the conversation and guide users to have a deeper communication. By putting relevant conversations into prompts, the system can remind users of the content discussed before, or provide background information for the next conversation, making the conversation more fluent and meaningful.

[0036] Specifically, to determine whether the first user's response has met the requirements of the opening remarks of the entire session, the following aspects can be used for evaluation:

[0037] Compare the user's response to the content of the opening statement to see if there is a clear correlation. If the user's response is highly correlated with the content of the prompt, then it can be assumed that the user has understood the opening statement and responded accordingly.

[0038] Analyze whether the user's response contains the keywords or phrases in the opening remarks. The presence of keywords can indicate the user's attention and understanding of the opening remarks.

[0039] Consider whether the user's response fits the context of the opening statement. Figure 1 If the user has understood the requirements of the opening remarks, then it can be assumed that the user has understood the requirements of the opening remarks.

[0040] Evaluate whether the user's response fully and clearly addresses the requirements of the opening statement. If the response is comprehensive and does not omit important information, then the user can be considered to have met the requirements.

[0041] Through natural language processing technology, semantic analysis is performed on user responses to determine whether they match the semantics of the opening remarks. This can help more accurately determine whether the user understands the requirements of the opening remarks.

[0042] Based on the evaluation results of the above aspects, it can be determined whether the first user's response has met the requirements of the opening remarks of the entire session.

[0043] S30: executing the process of the paragraph opening stage, which includes: providing the paragraph opening; obtaining the second user's reply corresponding to the paragraph opening; putting the dialogue related to the paragraph opening into the paragraph prompt words of the user's reply; judging whether the second user's reply has met the requirements of the paragraph prompt words; when it is determined that the second user's reply has met the requirements of the paragraph prompt words, further judging whether there are still necessary paragraphs or whether there are conditional paragraphs that meet the triggering conditions, if the judgment result is no, entering the end of the opening stage;

[0044] Specifically, when it is determined that the second user's reply has met the requirements of the paragraph prompt, it indicates that the user has made an appropriate response to the current paragraph content. Next, the system will further check whether there are necessary paragraphs or conditional paragraphs. Necessary paragraphs refer to certain key paragraphs that must be discussed or completed in the conversation. If the user's reply has met the requirements of the current paragraph prompt, but there are still unfinished necessary paragraphs, the system will guide the user to continue the discussion of these necessary paragraphs instead of entering the end of the opening remarks stage. Conditional paragraphs refer to paragraphs that will only appear according to specific conditions. The system will check whether there are conditional paragraphs that meet the conditions. If there are conditional paragraphs that meet the conditions, and the user's reply has met the requirements of the current paragraph prompt, the system will guide the user to enter the discussion of these conditional paragraphs. Only when the judgment result is "no", that is, there are no unfinished necessary paragraphs and conditional paragraphs that meet the conditions, the system will enter the end of the opening remarks stage. This shows that the user can smoothly end the conversation of the opening remarks and start the subsequent conversation content. In short, this process ensures the integrity and accuracy of the conversation, ensuring that all necessary and relevant paragraphs are properly discussed and completed before entering the next stage.

[0045] S40: executing a process of the end opening speech phase, which includes: providing an end opening speech; obtaining a third user reply corresponding to the end opening speech; putting the dialogue related to the end opening speech into the end opening speech prompt word of the user reply; judging whether the third user reply has met the end opening speech prompt word requirement; and providing a closing speech when the third user reply has met the end opening speech prompt word requirement;

[0046] S50: Propose a large model summary requirement, and generate a dialogue summary based on the large model summary requirement.

[0047] In a further embodiment, step S20 also includes the following steps: when it is determined that the user's reply does not meet the requirements of the full opening remarks prompt, further determine whether the full opening remarks related dialogue has reached the maximum round; if the maximum round has not been reached, give additional explanations or encouragement to the user; if the maximum round has been reached, enter the step of providing paragraph opening remarks in the paragraph opening remarks stage. Supplementary explanation means that the system will provide additional information or explanations so that the user can more clearly understand the meaning and background of the full opening remarks. This can include clarifying concepts, providing examples, or further describing related topics. Encouragement means that the system will give users positive feedback and encouragement to stimulate the user's interest and motivation to continue to participate in the dialogue. This can include affirming the user's efforts, expressing respect for the user's views, or encouraging the user to raise more questions and ideas.

[0048] In a further embodiment, step S30 also includes the following steps: if the paragraph prompt word requirement is not met, further determine whether the paragraph opening related dialogue has reached the maximum number of rounds; if the maximum number of rounds has not been reached, give additional instructions or encouragement to the user; if the maximum number of rounds has been reached, enter the step of determining whether there are still necessary paragraphs or whether there are conditional paragraphs that meet the trigger conditions.

[0049] In a further embodiment, step S40 also includes: determining whether the maximum number of rounds has been reached for the conversation related to the ending opening remarks; if the maximum number of rounds has not been reached for the conversation related to the ending opening remarks, providing additional instructions or encouragement to the user, and then entering the step of obtaining a reply from a third user in the current stage of ending the opening remarks; if the maximum number of rounds has been reached for the conversation related to the ending opening remarks, entering the step of providing a closing remark.

[0050] In a further embodiment, step S50 specifically includes: analyzing and summarizing the user's intentions, questions and emotional states to obtain a comprehensive understanding of the dialogue results. Specifically, the specific implementation may include the following steps: performing text analysis on the user's input, including identifying keywords, phrases, grammatical structures and semantic relationships, etc., to understand the user's intentions and the content of the expression. Analyzing the user's emotional state, the user's emotions, such as anger, happiness, sadness, etc., can be inferred through emotional vocabulary, tone, emoticons, etc. in the text. Combined with the user's intentions and emotional states, the dialogue results are summarized and summarized to obtain a comprehensive understanding. This can be presented to the user by generating natural language text or visual charts.

[0051] In a further embodiment, the method further includes, before step S10: setting a preset dialogue framework, the preset dialogue framework including: full opening remarks, paragraph opening remarks, ending opening remarks and closing remarks; the opening remarks of the dialogue framework can be dynamically replaced according to the user's personal information, which includes: pre-storing the user's personal information in a database; pre-setting an opening remarks template; obtaining the user's personal information from the database; replacing the obtained user personal information with the corresponding placeholders in the opening remarks template to generate personalized opening remarks; presenting the generated personalized opening remarks to the user; the placeholders in the opening remarks template are used to indicate the user's personal information that needs to be replaced.

[0052] In some embodiments, the dialogue background description is provided by the system role; the paragraph opening remarks and the closing remarks are provided by the assistant role; the closing remarks are provided by the assistant role to end the dialogue; and the model summary requirements are proposed by the system role to guide the dialogue model to summarize the dialogue. This technical solution has the following advantages: the system role and the assistant role each assume different responsibilities, making the information provision and task execution in the dialogue process clearer and more specific. The dialogue background description provided by the system role can provide the necessary contextual information for the dialogue model, which helps the model to better understand the content and background of the dialogue. The paragraph opening remarks and the closing remarks provided by the assistant role can guide the dialogue, making the dialogue more natural and smooth. The closing remarks provided by the assistant role can provide a clear ending signal for the dialogue, making the end of the dialogue clearer and more specific. The model summary requirements proposed by the system role can guide the dialogue model to summarize the dialogue, help extract key information and key points in the dialogue, and provide users with a more comprehensive and accurate dialogue summary.

[0053] In a further embodiment, the method further comprises the following steps: in the process of the conversation, the user's reply is packaged with a preset prompt word and explained to the conversation model; the content of the prompt word is dynamically adjusted according to the progress of the conversation and the user's reply; when the conversation model gives a corresponding answer and enters the next round of conversation, the prompt word of the previous round of conversation is deleted, and the new prompt word is used to package the user's reply in the new round. The advantage of this technical solution is that the method has the following advantages: the user's reply can be guided in a targeted manner by using the prompt word, making the conversation more focused and efficient. The prompt word can provide the conversation model with additional information about the progress and content of the conversation, which helps the model to better understand the conversation context. By dynamically adjusting the content of the prompt word, the method can adapt to different stages and needs of the conversation, making the conversation more natural and smooth. Deleting the prompt word of the previous round of conversation and using the new prompt word on the new round of user replies can avoid confusion and accumulation of the prompt word in the conversation, and ensure the clarity and coherence of the conversation. Through the reasonable use and adjustment of the prompt word, users can be guided to provide more valuable information, thereby improving the efficiency and accuracy of the conversation.

[0054] In a further embodiment, preset prompt words include: user information, conversation topic, current conversation stage, user address, supplement and prompt of big model information in the current conversation stage, analysis method and judgment criteria for user replies, encouragement to users when their replies cannot meet the judgment criteria, and structured reply restrictions that constrain the return form of the big model; the judgment criteria include: requirements for opening remarks prompt words for the entire conversation, requirements for paragraph prompt words, or requirements for ending opening remarks prompt words.

[0055] Specifically, the method of analyzing the user's answer content means that the system will use specific methods or standards to analyze the user's answer content. This analysis can help the system understand the user's intentions and needs more accurately, so as to make a more appropriate response. Specifically, the analysis method may include semantic understanding of the conversation content, sentiment analysis, keyword extraction, etc., to determine whether the user's answer is related to the topic of the conversation, whether it meets the preset judgment criteria, etc. In this way, the system can more effectively guide the conversation and improve the user experience.

[0056] Figure 3 is a functional block diagram of a dialogue control system based on an artificial intelligence dialogue model according to an embodiment of the present invention. Figure 3 As shown, the system 200 includes:

[0057] The dialogue background setting module 210 is used to provide a dialogue background description for the dialogue model to set the dialogue scene;

[0058] The full opening speech execution module 220 is used to execute the process of the full opening speech stage, including providing the full opening speech, obtaining the first user's reply corresponding to the full opening speech, putting the dialogue related to the full opening speech into the full opening speech prompt words of the user's reply, judging whether the first user's reply has met the requirements of the full opening speech prompt words, and triggering the paragraph opening speech execution module when it is determined that the first user's reply has met the requirements of the full opening speech prompt words;

[0059] The paragraph opening statement execution module 230 is used to execute the process of the paragraph opening statement stage, including providing the paragraph opening statement, obtaining the second user's reply corresponding to the paragraph opening statement, putting the paragraph opening statement related dialogue into the paragraph prompt word of the user's reply, judging whether the second user's reply has met the requirements of the paragraph prompt word, and when it is determined that the second user's reply has met the requirements of the paragraph prompt word, further judging whether there are still necessary paragraphs or whether there are conditional paragraphs that meet the triggering conditions, and if the judgment result is no, triggering the end of the opening statement execution module;

[0060] The end remark execution module 240 is used to execute the process of the end remark stage, including providing the end remark, obtaining the third user's reply corresponding to the end remark, putting the dialogue related to the end remark into the end remark prompt word of the user's reply, judging whether the third user's reply has met the end remark prompt word requirement, and providing the end remark when the third user's reply has met the end remark prompt word requirement;

[0061] The summary generation module 250 is used to propose a large model summary requirement and generate a dialogue summary according to the large model summary requirement.

[0062] In some embodiments, the full opening speech execution module 220 further includes:

[0063] The turn judgment submodule is used to further judge whether the conversation related to the full opening speech has reached the maximum turn when it is determined that the user's reply does not meet the requirements of the full opening speech prompt words;

[0064] A supplementary explanation or encouragement submodule is used to provide supplementary explanation or encouragement to the user when the maximum number of rounds has not been reached;

[0065] The paragraph opening remark triggering submodule is used to trigger the step of providing the paragraph opening remark in the paragraph opening remark execution module when the maximum round has been reached.

[0066] In some embodiments, the paragraph opening statement execution module 230 further includes:

[0067] The turn judgment submodule is used to further judge whether the dialogue related to the paragraph opening speech has reached the maximum turn when the paragraph prompt word requirement is not met;

[0068] A supplementary explanation or encouragement submodule is used to provide supplementary explanation or encouragement to the user when the maximum number of rounds has not been reached;

[0069] The conditional judgment submodule is used to trigger the step of judging whether there are still necessary paragraphs or whether there are conditional paragraphs that meet the triggering conditions when the maximum round has been reached.

[0070] In some embodiments, the end opening speech execution module 240 further includes:

[0071] The turn judgment submodule is used to judge whether the maximum number of turns has been reached for the dialogue related to the end of the opening remarks;

[0072] A supplementary explanation or encouragement submodule is used to provide supplementary explanation or encouragement to the user when the maximum number of rounds of dialogue related to the end of the opening remarks has not been reached;

[0073] The submodule for providing closing remarks is used to provide closing remarks when the maximum number of rounds of dialogue related to the closing opening remarks has been reached.

[0074] In some embodiments, the summary generation module specifically includes: an analysis and summary submodule for analyzing and summarizing the user's intentions, questions, and emotional states.

[0075] In some embodiments, the system also includes: a preset dialogue framework setting module, which is used to set a preset dialogue framework, and the preset dialogue framework includes: full opening remarks, paragraph opening remarks, ending opening remarks and closing remarks; the module also includes: a user information storage submodule, an opening remarks template setting submodule, a user personal information acquisition submodule, a personalized opening remarks generation submodule and a personalized opening remarks presentation submodule.

[0076] In some embodiments, the dialogue background description is provided by the system role; the paragraph opening remarks and closing remarks are provided by the assistant role; the closing remarks are provided by the assistant role; and the model summary requirement is proposed by the system role.

[0077] In some embodiments, the system also includes: a prompt word packaging and explanation module, which is used to use preset prompt words to package user replies during the conversation and explain them to the conversation model; the module also includes a dynamic adjustment submodule and a deletion and addition submodule.

[0078] In some embodiments, preset prompt words include user information, conversation topic, current conversation stage, address to the user, supplement and prompt to the big model information at the current conversation stage, analysis method and judgment criteria for user replies, encouragement to users when their replies fail to meet the judgment criteria, and structured reply restrictions that constrain the return form of the big model; the judgment criteria include requirements for opening remarks prompt words for the entire conversation, requirements for paragraph prompt words, or requirements for ending opening remarks prompt words.

[0079] The above technical solution is described in more detail below.

[0080] This mechanism imitates the scene of people communicating with each other to collect information or needs. Generally, an expert faces a user and conducts a pre-planned guided dialogue to collect the user's information or needs. For example: a mental health expert interviews a user to understand the user's mental health status; a bank account manager talks with the user to understand the business the user needs to handle and the goal; a sales customer service representative communicates with the user by phone to promote the product and explain the product details.

[0081] The above scenario can be abstracted into three parts:

[0082] The first part is the opening remarks of the whole event, which are preset or prepared in advance by the expert, including: the identity or role of the expert, what he can do or achieve for the user, a general description of what he is going to say next, a detailed description of the subsequent communication or dialogue method, and the words to attract users to continue communicating and dialogue.

[0083] After the expert finishes the first part of the explanation, the user will raise his or her own questions or indicate that he or she has understood the whole thing. If the user raises a question, the expert will answer the user's question or indicate that the subsequent dialogue will answer the user's question. If the user indicates that he or she has understood the whole thing, the expert will directly enter the second part. These dialogues related to the opening remarks after the full opening remarks are called full opening remarks related dialogues.

[0084] The second part is generally a number of paragraphs with similar structures. The main goal is for experts to give a detailed introduction to the demand points, topics or stages that need to be communicated one by one, while users express their own questions and experts explain the questions. Each paragraph here can be broken down into two parts: the paragraph opening and the paragraph opening related dialogue. There are two types of paragraphs: required paragraphs, which are paragraphs that must be communicated no matter what; conditional paragraphs, which are paragraphs that are triggered or not triggered by the different reactions of users in other paragraphs (for example, triggered by default, but the user's special reaction will cause it not to be triggered, or not triggered by default, but the user's special reaction will cause it to be triggered). There is a sequence between different paragraphs, because communication generally needs to maintain logic, but some special scenarios may also be random between paragraphs. After completing the second part, enter the third part.

[0085] The third part is mainly for the expert to summarize the conversation, confirm the user's needs or information, or use the preset words to end the conversation politely. The third part of the conversation can be broken down into the end of the opening remarks, the end of the opening remarks related dialogue, and the closing remarks.

[0086] Here is an example of a sales pitch scenario:

[0087] Part I:

[0088] Expert: Hello, Miss. We are glad that you answered our call. Our bank is promoting credit cards recently with many discounts. As our high-quality customer, please allow me to introduce our discounts to you in detail. (Opening remarks)

[0089] User: What bank are you from? (Opening remarks and related dialogue)

[0090] Expert: XX Bank. (Opening remarks and related dialogue)

[0091] User: OK. (Opening remarks and related dialogue)

[0092] Part II:

[0093] Paragraph 1:

[0094] Expert: Our bank's recent credit card application procedures are simple and can be done online, making the application very convenient. (Paragraph 1 opening remarks)

[0095] User: Can I do it online? (Dialogue related to the opening statement of paragraph 1)

[0096] Expert: Yes, you can open the web page on your mobile phone. (Dialogue related to the opening remarks of paragraph 1)

[0097] User: How do I do this? (Dialogue related to the opening statement of paragraph 1)

[0098] Expert: Open XXX on your phone. (Dialogue related to the opening statement of paragraph 1)

[0099] Paragraph 2:

[0100] Expert: In addition to the convenience of handling, our bank has recently cooperated with a large number of shopping malls. You can enjoy a 15% discount when you use our credit card to spend in shopping malls. (Paragraph 2 opening remarks)

[0101] User: What are the specific shopping malls? (Dialogue related to the opening statement of paragraph 2)

[0102] Expert: Including the XX shopping mall near you. (Dialogue related to the opening remarks of paragraph 2)

[0103] Part III:

[0104] Expert: Are you interested in applying for our bank's credit card? (End of opening remarks)

[0105] User: No, sorry. (End of opening statement)

[0106] Expert: Thank you for your call and I wish you a happy life. (Conclusion)

[0107] It can be found that in all parts, the topic is guided by experts and starts with an opening statement of explanation or interpretation. After the opening statement, users start asking questions about topics related to the opening statement, and the experts answer them, which is the relevant dialogue part.

[0108] The above process is the most commonly used pattern in human daily life, but it is difficult for the generative big model to complete and control. Even if a detailed process description is given to the big model at the beginning of the conversation, the big model will still forget during the continuous conversation. At the same time, when the second part contains many paragraphs and logical relationships, it is difficult for the big model to remember and conduct an effective conversation.

[0109] Therefore, the preset chain prompt active consultation dialogue process control mechanism based on the artificial intelligence dialogue big model proposed in the embodiment of the present invention aims to use a specific framework to guide the big model so that the big model can play the role of an expert in the above-mentioned dialogue process.

[0110] Before we explain the mechanism in detail, we should first clarify a few concepts, starting with the role. In the current mainstream generative dialogue model, there are generally three roles: system role (system, explaining the background or making requirements to AI), assistant role AI (assistant), and user role (user).

[0111] In the mechanism of the embodiment of the present invention, the user's words are not sent directly to the dialogue model, but are packaged with prompts before being sent. The prompts here will explain in detail the progress and topic of the current dialogue and the response that the model should have, to ensure that the model can give a correct response, for example:

[0112] User’s response: How does it work?

[0113] prompt: The user's reply is ---{reply}---. The topic currently being discussed with the user is how to apply for a credit card. Please explain the user's question accordingly. Steps to apply for a credit card: You need to open the XX website, enter XX, and then click XX to apply for a credit card.

[0114] In this example, the content that will actually be sent to the user of the large model is:

[0115] User: The user's reply is --- How do I do it specifically? ---. The topic currently being discussed with the user is how to apply for a credit card. Please explain the user's question accordingly. Steps to apply for a credit card: You need to open the XX website, enter XX, and then click XX to apply for a credit card.

[0116] There is another way to implement it, which is to use the system role (system) to prompt the prompt word, because the system will use the system role as the main prompter, for example:

[0117] User’s response: How does it work?

[0118] prompt: The topic currently being discussed with the user is how to apply for a credit card. Please explain the user's question based on the content of the next sentence. Steps to apply for a credit card: You need to open the XX website, enter XX, and then click XX to apply for a credit card.

[0119] So in this example, the content that will actually be sent to the large model is:

[0120] System: The topic currently being discussed with the user is how to apply for a credit card. Please explain the user's question based on the content of the next sentence. Steps to apply for a credit card: You need to open the XX website, enter XX, and then click XX to apply for a credit card.

[0121] User: How does it work specifically?

[0122] The embodiment of the present invention uses preset prompts to package the user's answers and explain them to the big model to ensure that the big model has an appropriate response. The prompts here include: user information, conversation topic, current conversation stage, user address, supplement and prompt to the big model information at the current stage, analysis method and judgment criteria for the user's answer content, and encouragement method for the user when the user's answer fails to meet the judgment criteria.

[0123] What needs to be further explained here is the "analysis method and judgment criteria for the user's answer content." For example, in the second part, in order to determine that the dialogue in this paragraph has been completed, the big model needs to use this content to judge the user's dialogue and determine that the dialogue goal of the current paragraph has been completed and it can enter the next stage.

[0124] When the big model gives the corresponding answer, when entering the next round of dialogue, the prompt of the previous round of dialogue needs to be deleted, and the prompt is used to wrap the answer of the new round of users, such as the example given above:

[0125] User: The user's reply is --- How do I do it specifically? ---. The topic currently being discussed with the user is how to apply for a credit card. Please explain the user's question accordingly. Steps to apply for a credit card: You need to open the XX website, enter XX, and then click XX to apply for a credit card.

[0126] Big model reply:

[0127] assistant: Open XXX on your mobile phone.

[0128] When generating the next dialogue, you still need to put this dialogue into the generation, but the prompt words need to be removed from the user content. The dialogue here will become when input into the model:

[0129] User: How does it work specifically?

[0130] assistant: Open XXX on your mobile phone.

[0131] In addition, the embodiment of the present invention also needs to require the return form of the big model in the prompt. Generally, by default, the return of the big model is a sentence, but when you want to obtain multiple structured information, such a return is very inconvenient. For example, this embodiment requires the big model to complete the following tasks based on the user's reply: determine the intention to apply for a credit card, give a reply to the user, and use sentiment analysis to analyze the emotions hidden in the user's conversation. The returns of the three tasks here are different. The first return is a Boolean value, the second return is a text, and the third return is an option, selecting one from multiple human emotions. If this embodiment directly describes the three requirements without constraining and explaining the reply method of the big model, then the reply of the big model may be as follows:

[0132] "The user has no intention of applying for a credit card. My response is: Thank you for answering the call, goodbye. The user's emotions are calm."

[0133] It is undoubtedly difficult to use such a result, because the return of the large model needs to be constrained in the prompt. One way used here is the json format. The latest interface of chatgpt has been implemented. You can set the format (format) when calling to ensure that the return of the large model is in json format. However, the mainstream large model in China does not have this parameter configuration, so it is necessary to make detailed instructions in the prompt to make the large model return in the form required by this embodiment, for example:

[0134] prompt: The user's reply is ---{reply}---. The topic currently being discussed with the user is how to apply for a credit card. Steps to apply for a credit card: You need to open the XX website, enter XX, and then click XX to apply for a credit card.

[0135] You need to respond with the following JSON:

[0136]

[0137] In this way, a standardized and structured response can be obtained, which makes it easier to further explore and use the capabilities of the large model.

[0138] The embodiment of the present invention has the ability to replace content. For example, if the opening remarks contain a name for a user, the pre-stored portion of the opening remarks where the user name is to be placed can be replaced with the user name stored in the database, for example:

[0139] Opening remarks: Hello {user_name}, we are glad that you answered our call. Our bank is currently promoting credit cards with many discounts. As our high-quality customer, please allow me to introduce our discounts to you in detail.

[0140] After replacement:

[0141] Opening remarks: Hello, Ms. Li, we are glad that you answered our call. Our bank’s credit card is being promoted recently with many discounts. As our high-quality customer, please allow me to introduce our discounts to you in detail.

[0142] In addition to the name, all personal information of the user can be filled in in real time to ensure the targeting of the user. The information here includes but is not limited to: name, gender, age, things done, and hobbies.

[0143] After all the above basic capabilities and dialog abstract capabilities are clear, the design of the embodiment of the present invention will be described.

[0144] The embodiment of the present invention uses a framework to simulate this process, and the implementation methods are explained one by one starting from the first part. Figure 2 As shown, it includes the following steps:

[0145] Step S10: First, before the first part, the background of the entire conversation and related information need to be explained to the large model. The information here is given using the system role (system). This embodiment calls this part the conversation background description (system_begin_prompt), for example:

[0146] system_begin_prompt: You are now a credit card sales manager and need to have a one-on-one conversation with the customer, explain to the user the current benefits and discounts of our bank's credit cards, the application process and other information, and confirm the user's application intention.

[0147] Step S20: Execute the process of the first part.

[0148] The first part of the process includes: the opening remarks and the dialogues related to the opening remarks. The opening remarks are called assistant_begin_prologue, and the prompt words of all users in the dialogues related to the opening remarks are called user_begin_prologue_prompt. Step 20 specifically includes the following steps S21 to S25:

[0149] Step S21: The assistant character (assistant) presents the prologue (assistant_begin_prologue).

[0150] Step S22: Obtain the user's reply.

[0151] Step S23: Put the conversation related to the prologue (prompt reply) into the prompt word of the prologue (user_begin_prologue_prompt) replied by the user.

[0152] Step S24: Determine whether the user's reply has met the prompt requirement. If it has not met the prompt requirement, execute step S25; if it has met the prompt requirement, execute step S31.

[0153] Step S25: If the prompt word requirement is not met, further determine whether the entire prologue-related dialogue has reached the maximum round (begin_prologue_max_round). If the maximum round has not been reached, execute step S26. If the maximum round has been reached, execute step S31.

[0154] Step S26: When the maximum number of rounds has not been reached, provide additional explanations or encouragement to the user. After step S26, the process proceeds to step S22.

[0155] Step S30: Execute the second part of the process, which includes the following steps:

[0156] Step S31: The assistant character (assistant) generates a section prologue (assistant_section_prologue).

[0157] Step S32: Obtain user reply.

[0158] Step S33: The section prologue related dialogue (prompt reply) includes the section prompt words (user_section_prologue_prompt) of the user's reply.

[0159] Step S34: Determine whether the user's reply has met the paragraph prompt word requirement. If it has not met the paragraph prompt word requirement, execute step S36; if it has met the paragraph prompt word requirement, execute step S35.

[0160] Step S35: Determine whether there are any required paragraphs or conditional paragraphs that meet the conditions. If the determination result is no, execute step S41.

[0161] Step S36: Determine whether the paragraph prologue related dialogue has reached the maximum round (section_prologue_max_round). If it has not reached the maximum round, execute step S37; if it has reached the maximum round, execute step S35.

[0162] Step S36: Give additional explanations or encouragement to the user, and then go to step S32.

[0163] Step S40: Execute the third part of the process, which specifically includes the following steps:

[0164] S41: The assistant role (assistant) provides or generates an ending prologue (assistant_end_prologue).

[0165] S42: Get the user's reply.

[0166] S43: Put the end prologue related dialogue (prompt reply) into the end prologue prompt of the user reply (user_end_prologue_prompt).

[0167] S44: Determine whether the user's reply has reached the end of the opening statement prompt word requirement.

[0168] S45: When the user's response has reached the requirement of ending the opening remarks prompt, the assistant role (assistant) provides or generates an ending statement (assistant_end).

[0169] S46: Determine whether the end prologue related dialogue has reached a maximum round (end_prologue_max_round).

[0170] S47: If the maximum number of rounds has not been reached before the opening remarks-related dialogue is ended, additional explanations or encouragements are given to the user, and then the process proceeds to step S42.

[0171] S45: If the maximum number of rounds of ending the opening remarks related dialogue has been reached, the process proceeds to step S45 where the assistant role (assistant) provides or generates the ending remarks (assistant_end).

[0172] S50: The system role (system) generates or proposes a model summary request (system_end_prompt).

[0173] The following are specific instructions:

[0174] When a conversation is started with a user, the first sentence the user hears is "assistant_begin_prologue", followed by a reply from the user, which is referred to as "reply" in this embodiment. At this time, the big model needs to be called for the first time. The call is combined with the background description before the conversation. The information input to the big model in this embodiment is as follows:

[0175] system:system_begin_prompt

[0176] assistant:assistant_begin_prologue

[0177] user:user_begin_prologue_prompt[reply]

[0178] Here user_begin_prologue_prompt[reply] is an explanation. The user's reply (reply) needs to be placed in user_begin_prologue_prompt.

[0179] If there is a relevant dialogue here, the follow-up to the first part of the dialogue should be as follows:

[0180] system:system_begin_prompt

[0181] assistant:assistant_begin_prologue

[0182] user:reply

[0183] assistant:ai_reply

[0184] user:user_begin_prologue_prompt[reply]

[0185] The judgment criteria for the end of the first part need to be given in user_begin_prologue_prompt, so that the big model can determine whether to terminate the current conversation and enter the second part. However, in actual use, the big model may make mistakes in judgment or the user's continuous conversation may not meet the requirements, which will cause the related conversation to continue indefinitely. Therefore, this embodiment needs to set a maximum round (max_round), and when the round of the conversation related to the full prologue reaches the maximum round, it is forced to enter the second part. All subsequent related conversations actually have similar problems, so each related conversation needs to set a maximum round. In order to distinguish, this embodiment names the maximum round of the conversation related to the full prologue as begin_prologue_max_round.

[0186] Each section in the second part is similar, including: section prologue assistant_section_prologue and section prologue related dialogue user_section_prologue_prompt, and also a maximum round control section_prologue_max_round.

[0187] The third part is divided into assistant_end_prologue, user_end_prologue_prompt, end_prologue_max_round, and assistant_end. The third part is special in that this embodiment requires the big model to summarize the conversation after the conversation ends, such as giving the user's willingness or questions about applying for a credit card, so this embodiment will add a system role requirement system_end_prompt at the end.

[0188] For example:

[0189] Part I:

[0190] system:system_begin_prompt

[0191] assistant:assistant_begin_prologue

[0192] user:reply

[0193] assistant:ai_reply

[0194] user:reply

[0195] Part II:

[0196] Paragraph 1:

[0197] assistant:assistant_section_1_prologue

[0198] user:reply

[0199] assistant:ai_reply

[0200] Paragraph 2:

[0201] assistant:assistant_section_2_prologue

[0202] user:reply

[0203] assistant:ai_reply

[0204] Part III:

[0205] assistant:assistant_end_prologue

[0206] user:reply

[0207] assistant:assistant_end

[0208] system:system_end_prompt

[0209] assistant:ai_reply

[0210] The final system role (system) allows the large model to complete tasks such as summarizing conversations, making better use of the capabilities of the large model.

[0211] The framework of this embodiment includes the following contents that need to be preset:

[0212] Before the conversation:

[0213] system_begin_prompt: Dialogue background description (system).

[0214] Part I:

[0215] assistant_begin_prologue: opening remarks (assistant);

[0216] user_begin_prologue_prompt: the prompt word for the user's reply (user);

[0217] begin_prologue_max_round: The maximum number of rounds for the entire prologue.

[0218] Part II:

[0219] assistant_section_prologue: paragraph prologue (assistant);

[0220] user_section_prologue_prompt: the prompt word for the user's reply (user);

[0221] section_prologue_max_round: The maximum number of rounds for the paragraph prologue-related dialogue;

[0222] Required paragraphs or conditional paragraphs;

[0223] The order or preconditions of paragraphs (in the form of triplets).

[0224] Part III:

[0225] assistant_end_prologue: ends the prologue;

[0226] user_end_prologue_prompt: prompt for user response (user);

[0227] end_prologue_max_round: The maximum number of rounds to end the prologue-related dialogue;

[0228] assistant_end: Ending remarks.

[0229] After the conversation:

[0230] system_end_prompt: Instructions and requirements for the model’s summary of the conversation.

[0231] Here, this embodiment is illustrated with a simple example:

[0232] Before the conversation:

[0233] system_begin_prompt:

[0234] You are now a credit card sales manager and need

[0235] Have one-on-one conversations with customers, explain to them the current benefits and offers of our bank's credit cards, application procedures, and other information, and confirm their intention to apply.

[0236] Part I:

[0237] assistant_begin_prologue:

[0238] {user_name} Hello, we are glad that you answered our call. Our bank is currently promoting its credit cards with many discounts. As our high-quality customer, please allow me to introduce our discounts to you in detail.

[0239] user_begin_prologue_prompt:

[0240] The user's reply is ---{reply}---. Currently the user explains that we are promoting credit cards.

[0241] You need to respond with the following JSON:

[0242]

[0243] begin_prologue_max_round: 5

[0244] Part II:

[0245] Paragraph 1:

[0246] assistant_section_prologue:

[0247] Our bank’s recent credit card application procedures are simple and can be processed online, making the application process very convenient.

[0248] user_section_prologue_prompt:

[0249] The user's reply is ---{reply}---. The topic currently being discussed with the user is how to apply for a credit card. Steps to apply for a credit card: You need to open the XX website, enter XX, and then click XX to apply for a credit card.

[0250] You need to respond with the following JSON:

[0251]

[0252] section_prologue_max_round: 5

[0253] Required paragraph

[0254] Paragraph 2:

[0255] assistant_section_prologue:

[0256] In addition to convenient processing, our bank has recently cooperated with a large number of shopping malls. You can enjoy a 15% discount when spending in shopping malls using our credit cards.

[0257] user_section_prologue_prompt:

[0258] The user's reply is ---{reply}---. The topic currently being discussed with the user is discounts for using credit cards in shopping malls. The shopping malls we cooperate with are XXX, XXX, and XXX.

[0259] You need to respond with the following JSON:

[0260]

[0261] section_prologue_max_round: 5

[0262] Required paragraph

[0263] Paragraph 3:

[0264] assistant_section_prologue:

[0265] You can tell me the location of your home and I will give you the location information of the nearest credit card processing branch.

[0266] user_section_prologue_prompt:

[0267] The user's reply is ---{reply}---. The topic currently being discussed with the user is the location of offline outlets for credit card processing. Our outlets are XXX, XXX, XXX.

[0268] You need to respond with the following JSON:

[0269]

[0270] section_prologue_max_round: 5

[0271] Conditional paragraph, when the user expresses that he will not use online processing.

[0272] Part III:

[0273] assistant_end_prologue:

[0274] Are you interested in applying for our credit card?

[0275] user_end_prologue_prompt:

[0276] The user's reply is ---{reply}---. Currently, we need to obtain the user's willingness to apply for a credit card.

[0277] You need to respond with the following JSON:

[0278]

[0279] end_prologue_max_round: 5

[0280] assistant_end:

[0281] Thank you for your call and wish you a happy life.

[0282] After the conversation:

[0283] system_end_prompt:

[0284] The conversation has ended. You need to reply in the following JSON format:

[0285]

[0286] Using this mechanism, after human experts preset the dialogue framework, the large model can play the role of an expert to communicate with users, greatly reducing manpower consumption. The standardized dialogue of the large model can also improve dialogue efficiency.

[0287] The beneficial technical effects of the technical solution of the embodiment of the present invention include:

[0288] Active dialogue guidance: This system can transform the artificial intelligence dialogue model from a passive information provider to an active dialogue guide, which is of great value in situations such as psychological counseling that require in-depth communication.

[0289] Adaptive multi-topic switching: By intelligently monitoring and analyzing user responses, the system can automatically switch between different preset topics or questions to achieve more targeted responses.

[0290] Conversation about sex.

[0291] Combination of professionalism and flexibility: By introducing opening remarks and questions designed by professionals, this system not only ensures the professionalism of the conversation, but also has a high degree of flexibility due to its adaptive and automatic switching characteristics.

[0292] Expansion of application areas: Although this process is designed with psychological counseling as an example, its proactive dialogue management characteristics make it potentially applicable to other situations that require professional guidance and multi-round, multi-topic communication.

[0293] In general, the embodiments of the present invention provide a new preset chain prompt active consultation dialogue process control mechanism based on the artificial intelligence dialogue model, which is expected to play an important role in psychological consultation, medical consultation and other professional consultation fields.

[0294] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0295] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, any one of the above methods is implemented.

[0296] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. Of course, there are other ways of readable storage media, such as quantum memory, graphene memory, etc. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media does not include electrical carrier signals and telecommunication signals.

[0297] The present invention also provides an electronic device. The electronic device of an embodiment of the present invention comprises: one or more processors; a storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the method provided by the present invention.

[0298] Reference below Figure 4 , which shows a schematic diagram of the structure of a computer system 800 of an electronic device suitable for implementing an embodiment of the present invention. Figure 4 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0299] like Figure 4 As shown, the computer system 800 includes a central processing unit (CPU) 801, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 802 or a program loaded from a storage part 808 into a random access memory (RAM) 803. In the RAM 803, various programs and data required for the operation of the computer system 800 are also stored. The CPU 801, the ROM 802, and the RAM 803 are connected to each other via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0300] The following components are connected to the I / O interface 805: an input section 806 including a keyboard, a mouse, etc.; an output section 807 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 808 including a hard disk, etc.; and a communication section 809 including a network interface card such as a LAN card, a modem, etc. The communication section 809 performs communication processing via a network such as the Internet. A drive 810 is also connected to the I / O interface 805 as needed. A removable medium 811, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 810 as needed so that a computer program read therefrom is installed into the storage section 808 as needed.

[0301] In particular, according to the embodiments disclosed in the present invention, the process described in the main step diagram above can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program code for executing the method shown in the main step diagram. In the above embodiment, the computer program can be downloaded and installed from the network through the communication part 809, and / or installed from the removable medium 811. When the computer program is executed by the central processing unit 801, the above functions defined in the system of the present invention are executed.

[0302] It should be noted that the computer-readable medium shown in the present invention may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present invention, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer readable signal media may also be any computer readable medium other than computer readable storage media, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0303] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the above-mentioned module, program segment or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0304] The units involved in the embodiments of the present invention may be implemented by software or hardware. The units described may also be provided in a processor, for example, may be described as: a processor including a pre-response unit, a receiving unit, and a request unit. The names of these units do not, in some cases, constitute limitations on the units themselves, for example, the pre-response unit may also be described as a "unit that provides a withdrawal amount and a withdrawal path to a request unit".

[0305] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may occur depending on design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A dialogue control method based on an artificial intelligence dialogue model, characterized in that: The following steps are involved: S10: Provides a dialogue background description for the dialogue model to set the dialogue scene; S20: executing the process of the full opening speech stage, which includes: providing the full opening speech; obtaining the first user's reply corresponding to the full opening speech; putting the dialogue related to the full opening speech into the full opening speech prompt words of the user's reply; judging whether the first user's reply has met the requirements of the full opening speech prompt words; when it is determined that the first user's reply has met the requirements of the full opening speech prompt words, entering the paragraph opening speech stage; S30: Execute the process of the paragraph opening stage, which includes: providing the paragraph opening; obtaining the second user's reply corresponding to the paragraph opening; putting the dialogue related to the paragraph opening into the paragraph prompt words of the user's reply; judging whether the second user's reply has met the requirements of the paragraph prompt words; when it is determined that the second user's reply has met the requirements of the paragraph prompt words, further judging whether there are still necessary paragraphs or whether there are conditional paragraphs that meet the triggering conditions, if the judgment result is no, entering the end of the opening stage; S40: executing the process of the end opening speech stage, which includes: providing the end opening speech; obtaining the third user's reply corresponding to the end opening speech; putting the dialogue related to the end opening speech into the end opening speech prompt word of the user's reply; judging whether the third user's reply has met the end opening speech prompt word requirement; when the third user's reply has met the end opening speech prompt word requirement, providing the end speech; S50: Propose a large model summary requirement, and generate a dialogue summary according to the large model summary requirement.

2. The method according to claim 1, characterized in that Step S20 also includes: When it is determined that the user's response does not meet the requirements of the opening speech prompt, further determining whether the conversation related to the opening speech has reached the maximum number of rounds; If the maximum number of rounds is not reached, additional instructions or encouragement are given to the user; If the maximum number of rounds has been reached, the process proceeds to the step of providing a paragraph opening statement in the paragraph opening statement phase.

3. The method according to claim 1, characterized in that Step S30 also includes: If the paragraph prompt word requirement is not met, further determine whether the dialogue related to the paragraph opening statement has reached the maximum number of turns; If the maximum number of rounds is not reached, additional instructions or encouragement to the user are given; If the maximum number of rounds has been reached, the process proceeds to the step of determining whether there are still required paragraphs or whether there are conditional paragraphs that meet the triggering conditions.

4. The method according to claim 1, characterized in that Step S40 also includes: Determine whether the maximum number of rounds of dialogue related to the end of the opening remarks has been reached; If the conversation related to the closing opening remarks has not reached the maximum number of rounds, then provide additional instructions or encouragement to the user, and then proceed to the step of obtaining a third user's reply in the closing opening remarks stage; If the maximum number of rounds of dialogue related to the closing opening remarks has been reached, the step of providing the closing remarks is entered.

5. The method according to claim 1, characterized in that Step S50 specifically includes: Analyze and summarize user intentions, questions, and emotional states to gain a comprehensive understanding of conversation outcomes.

6. The method according to claim 1, characterized in that Before step S10, the method further includes: A preset dialogue framework is set, and the preset dialogue framework includes: full opening remarks, paragraph opening remarks, closing opening remarks and closing remarks; the opening remarks of the dialogue framework can be dynamically replaced according to the user's personal information, which includes: pre-storing the user's personal information in a database; pre-setting an opening remarks template; obtaining the user's personal information from the database; replacing the obtained user personal information with the corresponding placeholder in the opening remarks template to generate a personalized opening remarks; presenting the generated personalized opening remarks to the user; the placeholder in the opening remarks template is used to indicate the user's personal information that needs to be replaced.

7. The method according to claim 1, characterized in that The dialogue background description is provided by the system role; The paragraph opening remarks and the closing opening remarks are provided by an assistant role; The ending words are provided by the assistant role and are used to end the conversation; The model summary requirement is proposed by the system role and is used to guide the large dialogue model to perform dialogue summary.

8. The method according to claim 1, characterized in that: The method further comprises: During the conversation, the user replies are packaged with preset prompt words and explained to the conversation model. The content of the prompt words is dynamically adjusted according to the progress of the conversation and the user's replies. When the conversation model gives the corresponding answer and enters the next round of conversation, the prompt words of the previous round of conversation are deleted, and the new prompt words are used to package the new round of user replies.

9. The method according to claim 8, characterized in that The preset prompt words include: user information, conversation topic, current conversation stage, user address, supplement and prompt of big model information in the current conversation stage, analysis method and judgment criteria for user replies, encouragement to users when their replies cannot meet the judgment criteria, and structured reply restrictions that constrain the return form of the big model; the judgment criteria include: requirements for opening remarks prompt words for the entire conversation, requirements for paragraph prompt words, or requirements for ending opening remarks prompt words.

10. A dialogue control system based on an artificial intelligence dialogue model, characterized in that: The system includes: The dialogue background setting module is used to provide dialogue background description for the dialogue model to set the dialogue scene; The full opening speech execution module is used to execute the process of the full opening speech stage, including providing the full opening speech, obtaining the first user's reply corresponding to the full opening speech, putting the dialogue related to the full opening speech into the full opening speech prompt words of the user's reply, judging whether the first user's reply has met the requirements of the full opening speech prompt words, and triggering the paragraph opening speech execution module when it is determined that the first user's reply has met the requirements of the full opening speech prompt words; A paragraph opening statement execution module is used to execute the process of the paragraph opening statement stage, including providing a paragraph opening statement, obtaining a second user reply corresponding to the paragraph opening statement, putting the paragraph opening statement related dialogue into the paragraph prompt word of the user reply, judging whether the second user reply has met the paragraph prompt word requirement, and when it is determined that the second user reply has met the paragraph prompt word requirement, further judging whether there are still necessary paragraphs or whether there are conditional paragraphs that meet the triggering condition, and if the judgment result is no, triggering the end of the opening statement execution module; an end remark execution module, used to execute the process of the end remark stage, including providing the end remark, obtaining the third user's reply corresponding to the end remark, putting the dialogue related to the end remark into the end remark prompt word of the user's reply, judging whether the third user's reply has met the end remark prompt word requirement, and providing the end remark when the third user's reply has met the end remark prompt word requirement; The summary generation module is used to propose a large model summary requirement and generate a dialogue summary according to the large model summary requirement.

Images