Intelligent interaction method and device, electronic device and computer readable storage medium

Abstract

This application discloses an intelligent interaction method, device, electronic device, and computer-readable storage medium, aiming to achieve more accurate and secure intelligent interaction and effectively improve the matching efficiency of intelligent interaction users. The method is applied to the intelligent interaction server and includes: generating an authorization token corresponding to the intelligent interaction user based on the authorization information of the intelligent interaction user regarding the digital avatar; generating controllable dialogue parameters corresponding to the intelligent interaction user based on the profile information of the intelligent interaction user; using an orchestration engine, referring to the current social scene identifier, authorization token, and controllable dialogue parameters to generate a dialogue strategy corresponding to the intelligent interaction user; controlling the digital avatar corresponding to the intelligent interaction user to perform intelligent interaction by referring to candidate dialogues, wherein the candidate dialogues are generated by the orchestration engine calling a language generation model based on the dialogue strategy; calculating the intention confidence of the intelligent interaction user during the intelligent interaction process, and outputting a user intervention prompt when the intention confidence exceeds a preset threshold.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to an intelligent interaction method, device, electronic device, and computer-readable storage medium. Background Technology

[0002] In scenarios such as local interest-based social networking or business matchmaking, users need to invest a lot of time in browsing, greeting, continuous dialogue and filtering, which leads to problems such as high communication costs, information overload and low cold start efficiency. Existing automated chat solutions, such as automated chatbots, can conduct dialogues when users are absent, but they usually only focus on text generation itself and lack strategic control that is consistent with the user's communication style and a controllable "low-risk exploration" mechanism. This can easily lead to problems such as privacy leaks and inappropriate expression, resulting in low matching efficiency in the intelligent interaction process.

[0003] Therefore, how to achieve more accurate and secure intelligent interaction in order to effectively improve the matching efficiency of intelligent interaction users and thus enhance the user experience is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0004] The purpose of this application is to provide an intelligent interaction method that can achieve more accurate and secure intelligent interaction, thereby effectively improving the matching efficiency of intelligent interaction users and thus enhancing the user experience. Another purpose of this application is to provide an intelligent interaction device, electronic device, and computer-readable storage medium, all of which have the above-mentioned beneficial effects.

[0005] Firstly, this application discloses an intelligent interaction method applied to an intelligent interaction server, comprising:

[0006] Obtain the authorization information of the intelligent interactive user regarding the digital clone, and generate an authorization token corresponding to the intelligent interactive user based on the authorization information;

[0007] Obtain the profile information of the intelligent interactive user, and generate controllable dialogue parameters corresponding to the intelligent interactive user based on the profile information;

[0008] Using an orchestration engine, a dialogue strategy corresponding to the intelligent interactive user is generated by referring to the current social scene identifier, the authorization token, and the controllable dialogue parameters.

[0009] The digital avatar corresponding to the intelligent interactive user is controlled to perform intelligent interaction with reference to candidate dialogues, which are generated by the orchestration engine calling the language generation model based on the dialogue strategy.

[0010] During the intelligent interaction process, the intention confidence level of the intelligent interaction user is calculated, and a user intervention prompt is output when the intention confidence level exceeds a preset threshold.

[0011] Optionally, the intelligent interaction method further includes:

[0012] Determine the sensitive triggering conditions corresponding to the current social scene identifier;

[0013] When sensitive information that triggers the aforementioned sensitive conditions is detected during intelligent interaction, a user intervention prompt is output.

[0014] Optionally, after outputting the user intervention prompt, the following may also be included:

[0015] When a permission modification instruction is received from the intelligent interactive user, the permission information is modified according to the permission modification instruction, and a determination is made based on the modification result of the permission information whether to continue to control the digital clone corresponding to the intelligent interactive user to perform intelligent interaction.

[0016] Alternatively, upon receiving a takeover command from the intelligent interactive user, the digital clone corresponding to the intelligent interactive user can be controlled to stop intelligent interaction.

[0017] Optionally, the intelligent interaction method further includes:

[0018] Acquire intelligent interaction records during intelligent interaction;

[0019] The intelligent interaction record is de-identified to obtain a de-identified interaction record;

[0020] Save the de-identified interaction records to a log file.

[0021] Optionally, the intelligent interaction method further includes:

[0022] When a revocation command is received regarding the authorization token, the digital clone corresponding to the smart interactive user is controlled to stop performing smart interaction;

[0023] The intelligent interaction records and / or the de-identified interaction records are deleted.

[0024] Alternatively, an unavailable flag can be added to the smart interaction record and / or the desensitized interaction record;

[0025] Alternatively, the intelligent interaction records and / or the desensitized interaction records may be encrypted and isolated.

[0026] Optionally, obtaining the profile information of the intelligent interactive user includes:

[0027] Obtain the historical interaction data and questionnaire survey data of the intelligent interactive users;

[0028] Based on the historical interaction data and the questionnaire survey data, the profile information of the intelligent interaction user is extracted.

[0029] Optionally, controllable dialogue parameters corresponding to the intelligent interactive user are generated based on the profile information, including:

[0030] Extract the psychological preference parameters of the intelligent interactive user based on the profile information;

[0031] Based on the mapping relationship between psychological preference parameters and controllable dialogue parameters, controllable dialogue parameters that match the psychological preference parameters corresponding to the intelligent interactive user are determined and used as the controllable dialogue parameters corresponding to the intelligent interactive user.

[0032] Secondly, this application discloses an intelligent interaction device applied to an intelligent interaction server, comprising:

[0033] The first generation module is used to obtain the authorization information of the intelligent interactive user regarding the digital clone, and generate an authorization token corresponding to the intelligent interactive user based on the authorization information;

[0034] The second generation module is used to obtain the profile information of the intelligent interactive user and generate controllable dialogue parameters corresponding to the intelligent interactive user based on the profile information.

[0035] The third generation module is used to generate a dialogue strategy corresponding to the intelligent interactive user by using the orchestration engine and referring to the current social scene identifier, the authorization token, and the controllable dialogue parameters.

[0036] The intelligent interaction module is used to control the digital clone corresponding to the intelligent interaction user to perform intelligent interaction with the candidate dialogues, which are generated by the orchestration engine calling the language generation model based on the dialogue strategy.

[0037] The prompt output module is used to calculate the intention confidence level of the intelligent interaction user during the intelligent interaction process, and output a user intervention prompt when the intention confidence level exceeds a preset threshold.

[0038] Thirdly, this application discloses an electronic device, including:

[0039] Memory, used to store computer programs;

[0040] A processor for executing the computer program to implement any of the intelligent interaction methods described above.

[0041] Fourthly, this application discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of any of the intelligent interaction methods described above.

[0042] This application provides an intelligent interaction method applied to an intelligent interaction server, comprising: obtaining authorization information of an intelligent interaction user regarding a digital avatar, and generating an authorization token corresponding to the intelligent interaction user based on the authorization information; obtaining profile information of the intelligent interaction user, and generating controllable dialogue parameters corresponding to the intelligent interaction user based on the profile information; using an orchestration engine, generating a dialogue strategy corresponding to the intelligent interaction user by referring to the current social scene identifier, the authorization token, and the controllable dialogue parameters; controlling the digital avatar corresponding to the intelligent interaction user to perform intelligent interaction by referring to candidate dialogues, wherein the candidate dialogues are generated by the orchestration engine calling a language generation model based on the dialogue strategy; calculating the intention confidence level of the intelligent interaction user during the intelligent interaction process, and outputting a user intervention prompt when the intention confidence level exceeds a preset threshold.

[0043] Applying the technical solution provided in this application, a technical solution for implementing intelligent interaction based on an intelligent interaction server is proposed. First, digital avatars are assigned to each participating intelligent interaction user. Each user then authorizes their digital avatar according to their needs, allowing the intelligent interaction server to generate an authorization token for each user based on the authorization information. Simultaneously, the intelligent interaction server can generate controllable dialogue parameters for controlling the dialogue style based on the user's profile information. Therefore, the intelligent interaction server can use an orchestration engine to generate dialogue strategies for each user based on the current social scene identifier, authorization token, and controllable dialogue parameters. It then calls a language generation model to generate candidate dialogues for each user during the intelligent interaction process. The server then controls the digital avatars of each user to implement intelligent interaction by referring to these candidate dialogues. This achieves intelligent interaction that adapts to user style (controllable dialogue parameters) and meets user security requirements (authorization token), resulting in more accurate and secure intelligent interaction. This effectively improves the matching efficiency of intelligent interaction users, thereby enhancing the user experience. In addition, this technical solution also calculates the intention confidence level of the intelligent interaction user during the intelligent interaction process. When the intention confidence level exceeds a preset threshold, a user intervention prompt is output to remind the corresponding intelligent interaction user to intervene manually, which can further improve the security performance during the intelligent interaction process.

[0044] The intelligent interactive device, electronic device, and computer-readable storage medium provided in this application also have the above-mentioned technical effects, and will not be described in detail here. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the prior art and the embodiments of this application, the accompanying drawings used in the description of the prior art and the embodiments of this application will be briefly introduced below. Of course, the accompanying drawings described below with respect to the embodiments of this application are only a part of the embodiments in this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort, and such other drawings also fall within the protection scope of this application.

[0046] Figure 1 A flowchart illustrating an intelligent interaction method provided in an embodiment of this application;

[0047] Figure 2 This is a schematic diagram of the structure of an intelligent interactive device provided in an embodiment of this application;

[0048] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0049] The core of this application is to provide an intelligent interaction method that can achieve more accurate and secure intelligent interaction, thereby effectively improving the matching efficiency of intelligent interaction users and thus enhancing the user experience. Another core aspect of this application is to provide an intelligent interaction device, electronic device, and computer-readable storage medium, all of which have the aforementioned beneficial effects.

[0050] To provide a clearer and more complete description of the technical solutions in the embodiments of this application, the technical solutions in the embodiments of this application will be described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0051] This application provides an intelligent interaction method.

[0052] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating an intelligent interaction method provided in an embodiment of this application. The intelligent interaction method is applied to an intelligent interaction server and may include, but is not limited to, the following S101~S105.

[0053] S101: Obtain the authorization information of the intelligent interactive user regarding the digital clone, and generate the authorization token corresponding to the intelligent interactive user based on the authorization information.

[0054] This step aims to generate authorization tokens for intelligent interaction users, which are the user objects participating in the intelligent interaction, and there must be at least two of them. The authorization tokens are generated based on the authorization information granted by the intelligent interaction user to their own digital avatar. The digital avatar is an intelligent agent representing the user in specific social scenarios, assisting in communication and decision-making, and can automatically converse within the scope of the user's authorization.

[0055] Specifically, a digital clone can be assigned to each smart interactive user, such as by installing it on the user terminal corresponding to the smart interactive user. The smart interactive user can set authorization information for their digital clone through the user terminal according to their actual needs and send it to the smart interactive server so that the smart interactive server can generate the corresponding authorization token for the smart interactive user based on the authorization information.

[0056] It is understandable that different smart interaction users can set different or the same authorization information. This authorization information can specifically be an authorization level, with different authorization levels corresponding to different authorization permissions (i.e., the permission information described below). In one possible implementation, the authorization levels can include four levels: Level L0—disclosing only anonymous interest tags and / or city-level information; Level L1—disclosing verifiable interest profile summaries and non-unique attributes (such as "weekend hiking / coffee / photography"); Level L2—disclosing more specific but still non-sensitive information (such as available appointment times, activity preferences, and regional locations); and Level L3—disclosing contact information / precise address / real identity identifier. To effectively ensure the security of smart interaction, when the smart interaction user has not set authorization information for their digital avatar, Level L0 can be set as the default disclosure level; when more specific but still non-sensitive information is needed, it can be upgraded to Level L2 within the user's authorization scope; and Level L3 requires user confirmation by default.

[0057] S102: Obtain the profile information of the intelligent interactive user, and generate controllable dialogue parameters corresponding to the intelligent interactive user based on the profile information.

[0058] This step aims to generate controllable dialogue parameters for intelligent interactive users. These controllable dialogue parameters are used to control the dialogue style of the corresponding intelligent interactive users and can be generated based on the profile information of the corresponding intelligent interactive users.

[0059] Obtaining user profile information for intelligent interaction can include: acquiring historical interaction data and questionnaire data of intelligent interaction users; and extracting user profile information based on historical interaction data and questionnaire data. In other words, user profile information can be obtained through questionnaires, extracted from historical interaction data, or both. To ensure accuracy, when using both methods simultaneously, weights can be assigned based on the specific circumstances. Furthermore, user profile information can also be obtained through user editing behavior, including but not limited to modifications, rewrites, selections, and rejections of candidate statements (described later).

[0060] The process of generating controllable dialogue parameters corresponding to intelligent interactive users based on profile information may include: extracting psychological preference parameters of intelligent interactive users based on profile information; and determining controllable dialogue parameters that match the psychological preference parameters of intelligent interactive users based on the mapping relationship between psychological preference parameters and controllable dialogue parameters, and using these as controllable dialogue parameters corresponding to intelligent interactive users.

[0061] Specifically, a mapping relationship between psychological preference parameters and controllable dialogue parameters for a broad user base can be pre-established. This allows for the extraction of psychological preference parameters from user profiles and their mapping to corresponding controllable dialogue parameters, thus determining the controllable dialogue parameters for each user. Psychological preference parameters (psych_params) include context dependence (high / low context), directness preference, and risk aversion preference. Controllable dialogue parameters (style_params) include directness, warmth, formality, structure, detail, assertiveness, hedging, and risk aversion (0-1). Based on this, taking "context dependence" as an example, mapping psychological preference parameters to controllable discourse parameters can include: for high-context users, increasing the proportion of warmth and hedging and decreasing directness, making the structure biased towards "relationship first, then matter, and providing room for exit"; for low-context users, increasing directness and structure and decreasing the proportion of hedging, making the structure biased towards "conclusion first, action items clear".

[0062] S103: Utilizes the orchestration engine to generate dialogue strategies for intelligent interactive users by referring to the current social scene identifiers, authorization tokens, and controllable dialogue parameters.

[0063] This step aims to generate dialogue strategies for intelligent interactive users. It involves inputting the current social scene identifier, authorization token, and controllable dialogue parameters into the orchestration engine for processing. The output is the dialogue strategy corresponding to the intelligent interactive user. The current social scene identifier indicates the type of the current intelligent interaction scenario, such as shared interests, event invitations, exploring common topics, or business matchmaking.

[0064] S104: Control the digital clone corresponding to the intelligent interactive user to perform intelligent interaction with reference to candidate dialogues. The candidate dialogues are generated by the orchestration engine calling the language generation model based on the dialogue strategy.

[0065] This step aims to achieve intelligent interaction based on the digital avatars corresponding to the intelligent interactive users. Specifically, during the intelligent interaction process of each digital avatar, the orchestration engine can call a pre-created language generation model to generate various candidate dialogues based on the aforementioned dialogue strategy, allowing each digital avatar to refer to these candidate dialogues to achieve intelligent interaction. The orchestration engine can write the mapping relationship between the aforementioned psychological preference parameters and controllable utterance parameters into the cue word template and decoding parameters to effectively ensure that the output of the language generation model maintains a stable style across different rounds.

[0066] S105: Calculate the intention confidence level of the intelligent interaction user during the intelligent interaction process, and output a user intervention prompt when the intention confidence level exceeds a preset threshold.

[0067] This step aims to calculate the intention confidence level of each intelligent interaction user during the intelligent interaction process. This allows for the output of a user intervention prompt to the corresponding intelligent interaction user's terminal when the intention confidence level exceeds a preset threshold, thus reminding the user to intervene manually. The intention confidence level is a 0-1 score calculated based on dialogue signals, which may include, but are not limited to, shared interest matching degree, time window matching degree, geographical accessibility, dialogue response quality, and number of dialogue rounds.

[0068] Understandably, intent confidence can be used to drive state machine transitions through exploration, intent confirmation, and transition phases. For example, a phased state machine can be defined as follows: S0—Not Started (matching candidates have not yet been reached); S1—Exploration Phase (disclosing L0 / L1 level); S2—Intent Confirmation Phase (disclosing L1 / L2 level); S3—Transition Preparation (triggering condition occurs, entering the confirmation node); S4—User Takeover / Confirmation Completed (authorization to enter L2 or L3 level); S5—End (successful transition / failure termination / timeout). Correspondingly, threshold examples are as follows: When intent_score ≥ 0.65, enter the intent confirmation phase; when intent_score ≥ 0.80 and no sensitive conditions are triggered, it can be suggested to generate an "event invitation draft"; when sensitive triggering conditions such as contact_exchange (exchanging contact information) or offline_meet (offline meeting) occur, force entry into S3 level and require confirmation from the corresponding intelligent interaction user. To avoid jitter, the intention confidence level can be set to migrate only if the condition is met for N consecutive rounds, or a time window can be used to accumulate scores.

[0069] Based on this, the intelligent interaction method can further include: determining the sensitive trigger conditions corresponding to the current social scene identifier; and outputting a user intervention prompt when sensitive information that hits the sensitive trigger conditions is detected during the intelligent interaction process. That is, the trigger conditions for outputting the user intervention prompt can also include sensitive information (i.e., the aforementioned sensitive trigger conditions). Different social scenes can correspond to different sensitive information, which may include, but are not limited to, exchanging contact information (contact_exchange), offline meetings / precise locations (offline_meet), money / payment / contract / commitment terms (money_commitment), sensitive topics (sensitive_topic), requests exceeding authorized data fields (over_permission), etc. Based on this, once a sensitive trigger condition is detected, the user can be forced to enter the intention confirmation stage.

[0070] Furthermore, after outputting the user intervention prompt, it may also include: when receiving a permission modification instruction from the intelligent interaction user, modifying the permission information according to the permission modification instruction, and determining whether to continue controlling the digital clone corresponding to the intelligent interaction user to perform intelligent interaction based on the result of the permission information modification; or, when receiving a takeover instruction from the intelligent interaction user, controlling the digital clone corresponding to the intelligent interaction user to stop performing intelligent interaction.

[0071] Specifically, when a smart interactive user is prompted to require human intervention, the user can choose to "take over" or "refuse" based on their situation. If the user chooses to "take over," their corresponding digital avatar will stop interacting with the user, facilitating human intervention. If the user chooses to "refuse," their corresponding digital avatar will also stop interacting with the user and will not be able to initiate human intervention, thus closing the conversation. Furthermore, when a smart interactive user initiates human intervention, they can modify permission information based on the situation, such as upgrading or downgrading the disclosure permission level. Clearly, a downgrade command will stop the corresponding digital avatar from interacting with the user, while an upgrade command will allow the corresponding digital avatar to continue interacting with the user.

[0072] In addition, to achieve the above functions, in the process of human intervention by intelligent interactive users, the user terminal of the intelligent interactive user can display candidate dialogues, disclose the reasons for the level change, and list of points to be confirmed through a visual interface, and provide entry points for confirmation, modification, rejection and takeover, so as to facilitate the input of relevant instructions by the intelligent interactive user.

[0073] In one embodiment of this application, the intelligent interaction method may further include: acquiring intelligent interaction records during the intelligent interaction process; performing desensitization processing on the intelligent interaction records to obtain desensitized interaction records; and saving the desensitized interaction records to a log file. In other words, the intelligent interaction method provided in this application embodiment can also achieve desensitized storage of intelligent interaction records, ensuring user information security while facilitating user traceability when needed later. Furthermore, the intelligent interaction server can also perform desensitized storage of disclosure level change records, reasons for triggering manual intervention, user confirmation results, dialogue strategy version identifiers, model call records, etc., generated during the intelligent interaction process.

[0074] Furthermore, the intelligent interaction method may also include: when a revocation instruction for the authorization token is received, controlling the digital clone corresponding to the intelligent interaction user to stop intelligent interaction; deleting intelligent interaction records and / or desensitized interaction records; or adding an unusable mark to intelligent interaction records and / or desensitized interaction records; or encrypting and isolating intelligent interaction records and / or desensitized interaction records.

[0075] Specifically, the authorization token generated by the intelligent interaction server based on the authorization information can be a revocable token. When the intelligent interaction server receives a revocation instruction for the authorization token, it can control the digital clone corresponding to the intelligent interaction user to stop intelligent interaction. At the same time, in order to ensure user information security, it can also delete, mark as unusable, or encrypt and isolate intelligent interaction records and other related data generated during the intelligent interaction process (such as disclosure level change records, reasons for triggering manual intervention, user confirmation results, dialogue strategy version identifiers, model call records, etc.).

[0076] As can be seen, the intelligent interaction method provided in this application first assigns digital avatars to the intelligent interaction users participating in the intelligent interaction, and the intelligent interaction users authorize their respective digital avatars according to their own needs, so that the intelligent interaction server can generate an authorization token corresponding to the intelligent interaction user based on the authorization information. At the same time, the intelligent interaction server can also generate controllable dialogue parameters for controlling the dialogue style based on the profile information of the intelligent interaction user. Thus, the intelligent interaction server can use the orchestration engine to generate dialogue strategies corresponding to each intelligent interaction user with reference to the current social scene identifier, authorization token, and controllable dialogue parameters, and call the language generation model to generate candidate dialogues for each intelligent interaction user in the intelligent interaction process, and then control the digital avatars corresponding to each intelligent interaction user to implement intelligent interaction with reference to the candidate dialogues. That is, it realizes intelligent interaction that adapts to the user style (controllable dialogue parameters) and meets the user's security needs (authorization token), that is, it realizes more accurate and secure intelligent interaction, which can effectively improve the matching efficiency of intelligent interaction users, thereby improving the user experience. In addition, this technical solution also calculates the intention confidence level of the intelligent interaction user during the intelligent interaction process. When the intention confidence level exceeds a preset threshold, a user intervention prompt is output to remind the corresponding intelligent interaction user to intervene manually, which can further improve the security performance during the intelligent interaction process.

[0077] For example, an embodiment of this application provides an intelligent interaction method.

[0078] First, the intelligent interaction method provided in this application embodiment is implemented based on an intelligent interaction system, which may include:

[0079] 1. User terminal: Displays a summary of the cloned conversation, candidate replies, points to be confirmed, and operation buttons. A 3D digital human display module is optional.

[0080] 2. Clone Platform (Intelligent Interaction Server): Stores user accounts, authorization policies, connector status, dialogue records, user profile parameters, and audit logs, etc.

[0081] 3. User Profile Building Module: This module builds user profiles from multiple data sources, including initial questionnaires, user editing behavior, and historical communication samples, to extract psychological preference parameters.

[0082] 4. Orchestration Engine: Selects dialogue strategies, prompt word templates, models and decoding parameters based on scene identifiers and profile parameters, and calls the large model to generate candidate dialogues.

[0083] 5. Privacy and Risk Control Module: Enforces disclosure level constraints, sensitive trigger detection, and mandatory rules for confirmation nodes.

[0084] 6. Intention Assessment Module: Calculates intention confidence based on dialogue signals and drives state machine transitions.

[0085] 7. Audit module: Records logs such as disclosure level changes, trigger reasons, user confirmation results and sending results, and can be stored in anonymized form.

[0086] Furthermore, based on the aforementioned intelligent interaction system, the implementation process of the intelligent interaction method provided in this application embodiment (taking two intelligent interaction users participating in intelligent interaction as an example) may include:

[0087] 1. Receive authorization information from the first user and the second user for their respective digital clones, generate a revocable authorization token, and determine the privacy disclosure policy corresponding to the authorization.

[0088] 2. Obtain user profiles from both parties to extract psychological preference parameters, and map these parameters to controllable dialogue parameters.

[0089] 3. Based on social scene identifiers (such as local interests / event invitations / exploration of common topics / business matchmaking, etc.), the orchestration engine generates dialogue strategies. The dialogue strategies include at least the output structure template, the intensity of the push, the upper limit of the disclosure level, and the sensitive trigger rules, and call the large model to generate candidate dialogues.

[0090] 4. Under the constraint of disclosure level L0 or L1, the first digital clone initiates an exploratory dialogue with the second digital clone (such as asking questions about common interests, activity preferences, and confirming time windows).

[0091] 5. During the dialogue, the intention assessment module updates the intention confidence based on the dialogue signals and transitions through the state machine from the "exploration phase → intention confirmation phase → transition phase".

[0092] 6. When the confidence level of intent reaches the threshold or a sensitive triggering condition is detected, the risk control module triggers user confirmation or takeover, such as outputting candidate text, key points to be confirmed, and suggested next steps to the user.

[0093] 7. After user confirmation, disclosure level can be upgraded or transfer actions can be performed (such as exchanging contact information, initiating offline invitations, etc.). If the user refuses, disclosure level will be downgraded or the conversation will be terminated.

[0094] 8. Generate and store audit logs and perform anonymization; when a user revokes authorization, stop roaming and perform actions such as deletion, unusable marking, or encryption isolation on related data.

[0095] Finally, this application also proposes the following two examples of intelligent interaction scenarios in different social settings:

[0096] Example 1: Local interest-based social roaming.

[0097] (1) Matching: Generate candidate objects based on distance within the same city, common interest tags, and availability within a time window;

[0098] (2) Exploration: The digital avatars of both parties engage in dialogue at L0 / L1 to identify common interests and activity preferences;

[0099] (3) Intention: If the confidence level of intention increases, the digital clone generates an "invitation draft" (such as weekend hiking / coffee).

[0100] (4) Transfer: When it is necessary to exchange contact information or confirm the specific location, the user confirms and the user can take over or authorize the process to continue.

[0101] Example 2: Business Cooperation Roaming.

[0102] (1) The scene identifier is switched to "Business Cooperation / Resource Exchange / Project Matching";

[0103] (2) More conservative disclosure strategy: By default, the company name, quotation, and contract information are not disclosed;

[0104] (3) Stricter triggering conditions: Once a quote, contract, payment, or commitment clause appears, the user is forced to take over;

[0105] (4) The audit is more complete, recording the triggering reasons, strategy version and confirmation results.

[0106] Therefore, the intelligent interaction method provided in this application has the following technical advantages:

[0107] (1) Complete low-risk social exploration when users are offline, significantly improving matching efficiency;

[0108] (2) Reduce the risk of privacy breaches and expression through gradual disclosure and confirmation steps;

[0109] (3) Improve communication consistency and controllability through strategy control driven by psychological preferences;

[0110] (4) Enhance user trust and traceability through auditing and revocation mechanisms;

[0111] (5) It can be extended to business cooperation scenarios to reduce compliance and commitment risks with stricter strategies.

[0112] This application provides an intelligent interactive device.

[0113] Please refer to Figure 2 , Figure 2 This is a schematic diagram of the structure of an intelligent interactive device provided in an embodiment of this application. The intelligent interactive device is applied to an intelligent interactive server and may include:

[0114] The first generation module 1 is used to obtain the authorization information of the intelligent interactive user regarding the digital clone, and generate the authorization token corresponding to the intelligent interactive user based on the authorization information;

[0115] The second generation module 2 is used to obtain the profile information of the intelligent interactive user and generate controllable dialogue parameters corresponding to the intelligent interactive user based on the profile information.

[0116] The third generation module 3 is used to generate dialogue strategies for intelligent interactive users by using the orchestration engine and referring to the current social scene identifier, authorization token, and controllable speech parameters.

[0117] Intelligent interaction module 4 is used to control the digital clone corresponding to the intelligent interaction user to perform intelligent interaction with the candidate dialogues. The candidate dialogues are generated by the orchestration engine calling the language generation model based on the dialogue strategy.

[0118] The prompt output module 5 is used to calculate the intention confidence level of the intelligent interaction user during the intelligent interaction process, and output a user intervention prompt when the intention confidence level exceeds a preset threshold.

[0119] As can be seen, the intelligent interaction device provided in this application first assigns digital avatars to the intelligent interaction users participating in the intelligent interaction, and the intelligent interaction users authorize their respective digital avatars according to their own needs, so that the intelligent interaction server can generate an authorization token corresponding to the intelligent interaction user based on the authorization information. At the same time, the intelligent interaction server can also generate controllable dialogue parameters for controlling the dialogue style based on the profile information of the intelligent interaction user. Thus, the intelligent interaction server can use the orchestration engine to generate dialogue strategies corresponding to each intelligent interaction user by referring to the current social scene identifier, authorization token, and controllable dialogue parameters, and call the language generation model to generate candidate dialogues for each intelligent interaction user in the intelligent interaction process, and then control the digital avatars corresponding to each intelligent interaction user to implement intelligent interaction by referring to the candidate dialogues. That is, it realizes intelligent interaction that adapts to the user style (controllable dialogue parameters) and meets the user's security needs (authorization token), that is, it realizes more accurate and secure intelligent interaction, which can effectively improve the matching efficiency of intelligent interaction users, thereby improving the user experience. In addition, this technical solution also calculates the intention confidence level of the intelligent interaction user during the intelligent interaction process. When the intention confidence level exceeds a preset threshold, a user intervention prompt is output to remind the corresponding intelligent interaction user to intervene manually, which can further improve the security performance during the intelligent interaction process.

[0120] In one embodiment of this application, the above-mentioned prompt output module 5 can also be used to determine the sensitive triggering conditions corresponding to the current social scene identifier; when sensitive information that hits the sensitive triggering conditions is detected during the intelligent interaction process, a user intervention prompt is output.

[0121] In one embodiment of this application, the intelligent interaction device may further include a control module, which, after outputting a user intervention prompt, modifies the permission information according to the permission modification instruction received from the intelligent interaction user, and determines whether to continue controlling the digital clone corresponding to the intelligent interaction user to perform intelligent interaction based on the modification result of the permission information; or, when receiving a takeover instruction from the intelligent interaction user, controls the digital clone corresponding to the intelligent interaction user to stop performing intelligent interaction.

[0122] In one embodiment of this application, the intelligent interaction device may further include a storage module for acquiring intelligent interaction records during the intelligent interaction process; performing desensitization processing on the intelligent interaction records to obtain desensitized interaction records; and saving the desensitized interaction records to a log file.

[0123] In one embodiment of this application, the smart interaction device may further include a data processing module, configured to, upon receiving a revocation instruction regarding the authorization token, control the digital clone corresponding to the smart interaction user to stop performing smart interaction; delete the smart interaction records and / or desensitized interaction records; or, add an unusable marker to the smart interaction records and / or desensitized interaction records; or, encrypt and isolate the smart interaction records and / or desensitized interaction records.

[0124] In one embodiment of this application, the second generation module 2 described above can be specifically used to obtain historical interaction data and questionnaire survey data of intelligent interactive users; and extract profile information of intelligent interactive users based on historical interaction data and questionnaire survey data.

[0125] In one embodiment of this application, the second generation module 2 described above can be specifically used to extract the psychological preference parameters of the intelligent interactive user based on the profile information; and to determine the controllable dialogue parameters that match the psychological preference parameters corresponding to the intelligent interactive user based on the mapping relationship between the psychological preference parameters and the controllable dialogue parameters, and use them as the controllable dialogue parameters corresponding to the intelligent interactive user.

[0126] For a description of the apparatus provided in the embodiments of this application, please refer to the above method embodiments; further details will not be repeated here.

[0127] This application provides an electronic device.

[0128] Please refer to Figure 3 , Figure 3 This application provides a schematic diagram of the structure of an electronic device, which may include:

[0129] Memory 11 is used to store computer programs;

[0130] The processor 10 is configured to execute computer programs and implement steps of any of the intelligent interaction methods described above.

[0131] like Figure 3 The diagram shows the structural composition of an electronic device, which may include a processor 10, a memory 11, a communication interface 12, and a communication bus 13. The processor 10, memory 11, and communication interface 12 all communicate with each other through the communication bus 13.

[0132] In this embodiment, the processor 10 may be a central processing unit (CPU), an application-specific integrated circuit, a digital signal processor, a field-programmable gate array, or other programmable logic devices.

[0133] The processor 10 can call programs stored in the memory 11. Specifically, the processor 10 can execute operations in the embodiments of the intelligent interaction method.

[0134] The memory 11 is used to store one or more programs. The programs may include program code, which includes computer operation instructions. In this embodiment, the memory 11 stores at least a program for implementing the following functions:

[0135] The system acquires authorization information from the intelligent interaction user regarding the digital avatar and generates an authorization token corresponding to the intelligent interaction user based on this information. It also acquires the intelligent interaction user's profile information and generates controllable dialogue parameters based on this profile. Using an orchestration engine, it generates a dialogue strategy for the intelligent interaction user by referencing the current social scene identifier, authorization token, and controllable dialogue parameters. The system controls the digital avatar corresponding to the intelligent interaction user to perform intelligent interaction based on candidate dialogues, which are generated by the orchestration engine using a language generation model based on the dialogue strategy. During the intelligent interaction process, the system calculates the intelligent interaction user's intention confidence level and outputs a user intervention prompt when the intention confidence level exceeds a preset threshold.

[0136] In one possible implementation, the memory 11 may include a program storage area and a data storage area, wherein the program storage area may store the operating system and applications required for at least one function; and the data storage area may store data created during use.

[0137] In addition, memory 11 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device or other volatile solid-state storage device.

[0138] Communication interface 12 can be an interface for the communication module, used to connect with other devices or systems.

[0139] Of course, it should be noted that, Figure 3 The structure shown does not constitute a limitation on the electronic device in the embodiments of this application. In practical applications, the electronic device may include more than Figure 3 More or fewer components as shown, or combinations of certain components.

[0140] This application provides a computer-readable storage medium.

[0141] The computer-readable storage medium provided in this application embodiment stores a computer program, which, when executed by a processor, can implement the steps of any of the intelligent interaction methods described above.

[0142] The computer-readable storage medium may include various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0143] For a description of the computer-readable storage medium provided in this application, please refer to the above method embodiments; further details will not be repeated here.

[0144] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section.

[0145] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0146] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0147] The technical solutions provided in this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are only for the purpose of helping to understand the methods and core ideas of this application. It should be noted that those skilled in the art can make several improvements and modifications to this application without departing from the principles of this application, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. An intelligent interaction method, characterized in that, Applications in intelligent interaction servers include: Obtain the authorization information of the intelligent interactive user regarding the digital clone, and generate an authorization token corresponding to the intelligent interactive user based on the authorization information; Obtain the profile information of the intelligent interactive user, and generate controllable dialogue parameters corresponding to the intelligent interactive user based on the profile information; Using an orchestration engine, a dialogue strategy corresponding to the intelligent interactive user is generated by referring to the current social scene identifier, the authorization token, and the controllable dialogue parameters. The digital avatar corresponding to the intelligent interactive user is controlled to perform intelligent interaction with reference to candidate dialogues, which are generated by the orchestration engine calling the language generation model based on the dialogue strategy. During the intelligent interaction process, the intention confidence level of the intelligent interaction user is calculated, and a user intervention prompt is output when the intention confidence level exceeds a preset threshold.

2. The intelligent interaction method according to claim 1, characterized in that, Also includes: Determine the sensitive triggering conditions corresponding to the current social scene identifier; When sensitive information that triggers the aforementioned sensitive conditions is detected during intelligent interaction, a user intervention prompt is output.

3. The intelligent interaction method according to claim 1 or 2, characterized in that, After outputting the user intervention prompt, it also includes: When a permission modification instruction is received from the intelligent interactive user, the permission information is modified according to the permission modification instruction, and a determination is made based on the modification result of the permission information whether to continue to control the digital clone corresponding to the intelligent interactive user to perform intelligent interaction. Alternatively, upon receiving a takeover command from the intelligent interactive user, the digital clone corresponding to the intelligent interactive user can be controlled to stop intelligent interaction.

4. The intelligent interaction method according to claim 1, characterized in that, Also includes: Acquire intelligent interaction records during intelligent interaction; The intelligent interaction record is de-identified to obtain a de-identified interaction record; Save the de-identified interaction records to a log file.

5. The intelligent interaction method according to claim 4, characterized in that, Also includes: When a revocation command is received regarding the authorization token, the digital clone corresponding to the smart interactive user is controlled to stop performing smart interaction; The intelligent interaction records and / or the de-identified interaction records are deleted. Alternatively, an unavailable flag can be added to the smart interaction record and / or the desensitized interaction record; Alternatively, the intelligent interaction records and / or the desensitized interaction records may be encrypted and isolated.

6. The intelligent interaction method according to claim 1, characterized in that, Obtaining the profile information of the intelligent interactive user includes: Obtain the historical interaction data and questionnaire survey data of the intelligent interactive users; Based on the historical interaction data and the questionnaire survey data, the profile information of the intelligent interaction user is extracted.

7. The intelligent interaction method according to claim 1, characterized in that, Based on the profile information, controllable dialogue parameters corresponding to the intelligent interactive user are generated, including: Extract the psychological preference parameters of the intelligent interactive user based on the profile information; Based on the mapping relationship between psychological preference parameters and controllable dialogue parameters, controllable dialogue parameters that match the psychological preference parameters corresponding to the intelligent interactive user are determined and used as the controllable dialogue parameters corresponding to the intelligent interactive user.

8. An intelligent interactive device, characterized in that, Applications in intelligent interaction servers include: The first generation module is used to obtain the authorization information of the intelligent interactive user regarding the digital clone, and generate an authorization token corresponding to the intelligent interactive user based on the authorization information; The second generation module is used to obtain the profile information of the intelligent interactive user and generate controllable dialogue parameters corresponding to the intelligent interactive user based on the profile information. The third generation module is used to generate a dialogue strategy corresponding to the intelligent interactive user by using the orchestration engine and referring to the current social scene identifier, the authorization token, and the controllable dialogue parameters. The intelligent interaction module is used to control the digital clone corresponding to the intelligent interaction user to perform intelligent interaction with the candidate dialogues, which are generated by the orchestration engine calling the language generation model based on the dialogue strategy. The prompt output module is used to calculate the intention confidence level of the intelligent interaction user during the intelligent interaction process, and output a user intervention prompt when the intention confidence level exceeds a preset threshold.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor for implementing the steps of the intelligent interaction method as described in any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the intelligent interaction method as described in any one of claims 1 to 7.

Images