Intelligent agent generation method and apparatus, interaction method and apparatus, medium and device

Abstract

The present disclosure relates to an intelligent agent generation method and apparatus, an interaction method and apparatus, a medium and a device. The intelligent agent generation method comprises: displaying an interaction interface corresponding to an intelligent agent; on the basis of the interaction interface, obtaining a historical file of a target user corresponding to the intelligent agent, the historical file comprising an operation record of the target user in a historical time period; and on the basis of the historical file, obtaining a target intelligent agent corresponding to the target user. Therefore, data can be provided for an intelligent agent in a file interaction mode, data acquisition of a target intelligent agent is facilitated, and the operation complexity of generating the target intelligent agent is reduced; additionally, by generating a corresponding target intelligent agent by means of a historical file of a target user, the matching degree between the target intelligent agent and the target user performance can be improved, the consistency between the generated target intelligent agent and the user demand is improved, and the anthropomorphic level and the personalized level of the generated target intelligent agent are improved, so that the user can carry out personalized interaction with the target intelligent agent.

Description

Method, device, medium and equipment for generating an agent, and method and device for interacting with an agent

[0001] This application claims priority to Chinese Patent Application No. 202410702873.0, filed on May 31, 2024, the disclosure of which is incorporated herein in its entirety as part of the present application. TECHNICAL FIELD

[0002] The present disclosure relates to a method, device, medium, equipment and computer program product for generating an agent. BACKGROUND

[0003] With the development of computer technology, users can interact with chat robots to achieve communication. In the related art, chat robots can provide communication with users, which usually relies on general communication scripts and NLP (Natural Language Processing) capabilities, and it is difficult to achieve personalized interaction. SUMMARY

[0004] This summary is provided to introduce a selection of concepts, which are further described below in the detailed description. This summary is not intended to identify key or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter.

[0005] In a first aspect, the present disclosure provides a method for generating an agent, the method comprising:

[0006] displaying an interaction interface corresponding to the agent;

[0007] obtaining a history file of a target user corresponding to the agent based on the interaction interface, the history file containing operation records of the target user in a historical period;

[0008] obtaining a target agent corresponding to the target user based on the history file.

[0009] In a second aspect, the present disclosure provides an interaction method, the method comprising:

[0010] receiving input information of a user;

[0011] determining reply information corresponding to the input information based on a target agent, wherein the target agent is determined based on the method for generating an agent of the first aspect;

[0012] outputting the reply information based on an input type of the input information.

[0013] In a third aspect, the present disclosure provides an apparatus for generating an agent, the apparatus comprising:

[0014] a first display module configured to display an interaction interface corresponding to the agent;

[0015] a obtaining module configured to obtain a historical file of a target user corresponding to the agent based on the interaction interface, the historical file comprising operation records of the target user in a historical period;

[0016] a processing module configured to obtain a target agent corresponding to the target user based on the historical file.

[0017] In a fourth aspect, the present disclosure provides an interaction apparatus, the apparatus comprising:

[0018] a second receiving module configured to receive input information of a user;

[0019] a second determining module configured to determine reply information corresponding to the input information based on a target agent, wherein the target agent is determined based on the method for generating an agent according to the first aspect;

[0020] an output module configured to output the reply information based on an input type of the input information.

[0021] In a fifth aspect, the present disclosure provides a computer readable medium having a computer program stored thereon, wherein the computer program is executed by a processing apparatus to implement the steps of the method according to the first aspect or the second aspect.

[0022] In a sixth aspect, the present disclosure provides an electronic device comprising:

[0023] a storage apparatus having a computer program stored thereon;

[0024] a processing apparatus configured to execute the computer program in the storage apparatus to implement the steps of the method according to the first aspect or the second aspect.

[0025] In a seventh aspect, the present disclosure provides a computer program product comprising a computer program, wherein the computer program is executed by a processor to implement the steps of the method according to the first aspect or the second aspect.

[0026] Other features and advantages of the present disclosure will be illustrated in the following detailed description. BRIEF DESCRIPTION OF DRAWINGS

[0027] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and the originals and elements are not necessarily drawn to scale. In the drawings:

[0028] Figure 1 is a flowchart of a method for generating an intelligent agent according to some embodiments of the present disclosure;

[0029] Figure 2 is a schematic diagram of a target configuration interface provided according to some embodiments of the present disclosure;

[0030] Figure 3 is a flowchart of an interaction method provided according to some embodiments of the present disclosure;

[0031] Figure 4 is a UML diagram corresponding to the interaction method provided in the embodiments of this disclosure;

[0032] Figure 5 is a flowchart of text or voice interaction in a chat page provided based on an embodiment of the present disclosure;

[0033] Figure 6 is a flowchart of video interaction in a chat page provided based on an embodiment of the present disclosure;

[0034] Figure 7 is a block diagram of an agent generation apparatus according to some embodiments of the present disclosure;

[0035] Figure 8 is a block diagram of an interactive device provided according to some embodiments of the present disclosure; and

[0036] Figure 9 shows a schematic diagram of the structure of an electronic device suitable for implementing embodiments of the present disclosure. Detailed Implementation

[0037] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0038] It should be understood that the steps described in the method embodiments of this disclosure may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of this disclosure is not limited in this respect.

[0039] The term "comprising" and its variations as used herein are open-ended inclusions, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Definitions of other terms will be given in the description below.

[0040] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0041] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0042] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0043] It is understood that before using the technical solutions disclosed in the various embodiments of this disclosure, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this disclosure in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained.

[0044] For example, upon receiving a user's active request, a prompt message is sent to the user to explicitly inform them that the requested operation will require the acquisition and use of the user's personal information. This allows the user to independently choose whether to provide personal information to the software or hardware, such as the electronic device, application, server, or storage medium performing the operations of this disclosed technical solution, based on the prompt message.

[0045] As an optional but non-limiting implementation, in response to a user's active request, sending a prompt message to the user can be done via a pop-up window, where the prompt message can be presented in text format. Furthermore, the pop-up window can also include a selection control allowing the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0046] It is understood that the above notification and user authorization process are merely illustrative and do not constitute a limitation on the implementation of this disclosure. Other methods that comply with relevant laws and regulations may also be applied to the implementation of this disclosure.

[0047] Meanwhile, it is understood that the data involved in this technical solution (including but not limited to the data itself, the acquisition or use of the data) shall comply with the requirements of relevant laws, regulations and related provisions.

[0048] Figure 1 shows a flowchart of a method for generating an intelligent agent according to some embodiments of the present disclosure. As shown in Figure 1, the method may include:

[0049] In step 11, the interactive interface corresponding to the intelligent agent is displayed.

[0050] The interactive interface of the intelligent agent can include interfaces that interact with the agent's output. For example, the interface can include a configuration interface for the intelligent agent, allowing users to generate or adjust the agent through configuration operations. Alternatively, the interface can include a dialogue interface, allowing users to provide data to the agent through dialogue and subsequently adjust the agent based on that data.

[0051] In step 12, based on the interactive interface, the historical file of the target user corresponding to the intelligent agent is obtained, and the historical file contains the operation records of the target user within the historical time period.

[0052] The historical time period can be set and selected based on user needs. This operation record can be the target user's actions in multiple interactive applications. The historical file can be a file that the user has exported or obtained from other interactive applications beforehand. If the interactive application supports file export, the user can export the chat history file with the target user and the publicly posted content corresponding to the target user from that interactive application. As an example, when exporting chat history, text data, audio data, and video data can be exported separately, thus obtaining text files, audio files, and video files. If the interactive application does not support export, the target user's operation record can be stored in a file through editing. For example, for audio and video data, the user can manually copy and save it to the corresponding file to obtain the historical file. For example, for text data, the user can manually edit and copy, or take a screenshot and then use OCR (Optical Character Recognition) technology to recognize characters and obtain the corresponding text file.

[0053] In step 13, the target agent corresponding to the target user is obtained based on historical files.

[0054] As an example, the data corresponding to the historical file can be added to the agent's memory bank as the agent's memory data. This allows the agent to have memory data of the target user's operation records within a historical period, and the agent that added the historical file becomes the target agent for that target user. Accordingly, when engaging in dialogue with the target agent for the target user later, for example, by inputting the dialogue text "What are your favorite movies?", the target agent can simultaneously determine its existing memory file and the historical file during the output process, thereby generating a corresponding response.

[0055] As another example, this step can train the agent based on data from historical files to obtain a target agent corresponding to the target user. For instance, this step can implement the target agent based on an artificial intelligence model. For example, it can use machine learning and deep learning techniques commonly used in the field, with the data in the historical files as training data, to train an AI (Artificial Intelligence) model, thereby obtaining a target agent that simulates the target user and improving the anthropomorphism and personalization of the target agent.

[0056] Therefore, in the above technical solution, historical files corresponding to the target user can be obtained. These historical files are then provided to the intelligent agent via file interaction, and the target intelligent agent corresponding to the target user is obtained from these files. This technical solution allows for data provision to the intelligent agent through file interaction, facilitating data acquisition for the target intelligent agent and reducing the operational complexity of generating the target intelligent agent. Furthermore, generating the corresponding target intelligent agent from the target user's historical files improves the matching degree between the target intelligent agent and the target user's behavior, enhances the consistency between the generated target intelligent agent and user needs, and increases the anthropomorphism and personalization level of the generated target intelligent agent, enabling users to engage in personalized interactions with the target intelligent agent.

[0057] In one possible embodiment, the interactive interface is a target configuration interface for configuring the agent.

[0058] To enhance the user experience during interactions, users can configure the interactive agent based on their needs, thereby increasing the personalization of the generated agent. This agent can be a digital human created using AI technology. A digital human is a virtual object created by combining computer graphics, artificial intelligence, machine learning, speech synthesis, and other technologies to simulate the appearance, behavior, and interaction capabilities of a real human. Accordingly, in this step, the user can trigger a configuration operation to display the target configuration interface on the terminal.

[0059] Accordingly, an example implementation of obtaining the historical files of the target user corresponding to the intelligent agent based on the interactive interface may include:

[0060] In response to the selection operation of the historical data configuration item in the target configuration interface, the historical data upload interface is displayed;

[0061] The files received based on the data upload interface are used as the historical files.

[0062] As shown in Figure 2, the target configuration interface displays multiple configurable historical data configuration items. Each item is used to upload or import historical files of a specific type. In this step, users can upload the corresponding historical files by selecting the appropriate historical data configuration item. For example, selection can be made by clicking. If a user wants to upload chat history data, they can click on the corresponding historical data configuration item A2 to select it, which will then display the chat history historical data upload interface.

[0063] Accordingly, users can upload data in the historical data upload interface. For example, users can select files from their local machine or the cloud using a file selection method, and then submit the file. The historical data upload interface may also display file controls; clicking these controls will upload the corresponding file, and the selected file will be added to the historical file list in response to the file submission operation.

[0064] As another example, the interactive application corresponding to the target user may also have a public data acquisition interface. In this case, historical files can also be obtained through the interface. Then, after clicking the corresponding upload control in the historical data upload interface, the interface of the interactive application can be displayed. Based on the target user's verification information, the historical files corresponding to the target user can be directly obtained from the interactive application based on the data acquisition interface.

[0065] Therefore, through the above technical solution, during the configuration of the intelligent agent, the target user's historical files can be uploaded so that the intelligent agent can be trained directly based on the historical files, thereby obtaining the target intelligent agent corresponding to the target user. This facilitates the acquisition of training data for the target intelligent agent, reduces the complexity of the user's configuration of the target intelligent agent, provides more comprehensive and reliable data support for the training of the target intelligent agent, and ensures the consistency between the generated target intelligent agent and the user's needs.

[0066] As an example, the historical file contains at least one of audio data, text data, video data, and published content. Different types of data can be stored in the same file or in different files. For instance, taking the target user's chat history within a historical period as an example, text records in the chat history can be used as text data, voice messages sent in the chat history can be used as audio data, and videos sent in the chat history can be used as video data. As another example, the historical file may also contain the target user's published content within a historical period, i.e., information publicly released by the target user, such as likes, comments, or text and video posts on social media. Therefore, the historical file can be used to profile the target user's interactive behavior and characteristics from multiple dimensions. By integrating multi-source information about the target user, an accurate representation of the target user can be achieved, providing accurate and personalized data for the subsequent generation of intelligent agents.

[0067] This process involves exporting the chat history between the current user and the target user in the corresponding interactive application to obtain the corresponding chat history file. In this step, the chat history file can be uploaded through the historical data upload interface so that the target agent can be obtained based on the chat history file later, thereby improving its matching degree with the target user.

[0068] In one possible embodiment, the interactive interface is an interface for dialogue with the intelligent agent, where U1 is the user side and B1 is the intelligent agent side. User U1 can interact with B1 by inputting in the dialogue interface.

[0069] Accordingly, an exemplary implementation of obtaining the historical files of the target user corresponding to the intelligent agent based on the interactive interface may include:

[0070] During the dialogue process of the intelligent agent, the target user's historical files are received from the interactive interface.

[0071] In this embodiment, the user can directly input a file from the interactive interface. A file icon can be displayed in the dialog window, allowing the user to click the file icon to access the file interface, select the desired file, and upload it. The file can then be provided to the intelligent agent via a dialog. During this process, the user can input a description of the file in the dialog box, enabling the intelligent agent to perform corresponding operations based on the historical file. For example, the intelligent agent can add the historical file as memory data to its corresponding memory file based on the user's input description. Alternatively, the user can upload a file by dragging and dropping it into the dialog box, thus receiving the target user's historical files from the interactive interface.

[0072] Therefore, through the above technical solution, historical files can be uploaded to the intelligent agent during the interaction process, so as to adjust the intelligent agent to obtain the target intelligent agent, so that the target intelligent agent can contain the target user's historical operations and improve the personalized interaction of the target intelligent agent.

[0073] In one possible embodiment, an exemplary implementation of obtaining the target agent corresponding to the target user based on the historical files may include:

[0074] Based on the historical files, the historical memory data corresponding to the intelligent agent is determined, wherein the historical memory data is at least a portion of the data in the historical files;

[0075] Based on the historical memory data, the target intelligent agent corresponding to the target user is obtained.

[0076] As an example, data in uploaded historical files can be directly identified as historical memory data, and the type of historical memory data can be determined based on the historical data configuration items triggered when the historical file was uploaded. For instance, if a user triggers the display of the historical data upload interface by operating the chat history configuration items, the uploaded historical file can be used as the historical memory data corresponding to the chat history, so as to distinguish between various uploaded historical memory data and provide multiple types of training data for subsequent training of the agent.

[0077] As another example, the user can be shown data from the historical file, and then select a portion of it as historical memory data. For instance, if chat logs are displayed, the user can select chat logs from the most recent month as historical memory data, thus filtering the data in the historical file. Subsequently, based on the historical memory data, a target agent corresponding to the target user can be obtained. For example, the historical memory data can be added to the agent's memory bank to obtain the target agent, or it can be used as training data to train the agent and adjust its performance characteristics to obtain the target agent. The specific implementation methods for obtaining the target agent have been detailed above and will not be repeated here.

[0078] Therefore, the above technical solution allows for direct filtering of data in historical files, enabling users to select historical memory data provided by the intelligent agent, thus meeting user needs and improving user satisfaction with the target intelligent agent.

[0079] As another example, the method of determining the historical memory data corresponding to the agent based on the historical file may include:

[0080] The interface for processing historical files is displayed. Specifically, this interface can be displayed after the user selects a historical file. For example, for a text file, the interface can display the text content contained within that file, allowing the user to edit it as needed.

[0081] In one scenario: in response to receiving a user's editing operation on the historical file in the processing interface, the data in the file obtained after editing the historical file can be used as the historical memory data.

[0082] In practical applications, the content of directly exported text files may contain information that the user does not want to provide to the agent, or content that the user deems unsuitable for adjusting the agent. In this embodiment, the user can edit the content of the historical file, such as deleting, adjusting, or adding to the displayed text. The data in the file obtained after the user's editing of the historical file can then be used as the historical memory data.

[0083] In another scenario: in response to receiving a user's selection operation on the time selection control in the processing interface, a target time period can be determined, and the data within the target time period in the historical file can be identified as the historical memory data.

[0084] In one possible application scenario, when a user wants to adjust the intelligent agent corresponding to a target user, they may only want to use a portion of the data in the historical file. This disclosure provides a method for selecting content from the historical file through this embodiment. For example, the historical file may contain timestamp information of its data; correspondingly, the historical file processing interface may include a time selection control to quickly enable the user to select a portion of the content from the historical file.

[0085] As an example, if the historical file is a file generated from the target user's interactive operations within a historical time period, then when exporting the historical file, the timestamp information corresponding to the content data in the historical file can be obtained simultaneously. Accordingly, in this embodiment, the user can quickly select content from the historical file based on the timestamp information. For example, the time selection control in the processing interface can include a start time control and an end time control. The user can determine the time period they want to upload based on their own needs. For example, they can configure the start time based on the start time control and the end time based on the end time control, thereby determining the target time period based on the times corresponding to the two controls.

[0086] Then, by combining the timestamp information of the historical file with the target time period, the content within that target time period can be obtained. Taking a text file of chat logs as an example, if the target time period is from April 1, 2023 to April 1, 2024, the text content of the chat logs corresponding to April 1, 2023 to April 1, 2024 can be used as the historical memory data. As another example, the data in the historical file can also be adjusted by combining the filtering methods in the two scenarios mentioned above to obtain historical memory data.

[0087] Therefore, through the above technical solution, when uploading the target user's historical files, the user can filter the data to be uploaded from the historical files, and can also edit the data to accurately upload the data that the user determines needs, so as to ensure that the uploaded data can meet the user's usage needs, thereby improving the user's satisfaction with the generated target intelligent agent, ensuring that the target intelligent agent restores the characteristics of the target user, improving the personalization of the target intelligent agent, and enhancing the user's interaction experience with the target intelligent agent.

[0088] In one possible embodiment, an exemplary implementation of determining the historical memory data corresponding to the agent based on the historical file is as follows, and this step may include:

[0089] The historical file is parsed to obtain multiple sub-files corresponding to the historical file.

[0090] As an example, historical files can be parsed based on a content understanding model, which can be an intelligent agent that interacts with the user, or other understanding models in the field, such as pre-set models for content understanding, such as understanding models implemented by large language models, etc. This disclosure does not limit them.

[0091] Then, the file information of the multiple sub-files is displayed; in response to determining the sub-file selected by the user, the data of the selected sub-file is used as the historical memory data.

[0092] As an example, the interactive interface is a target configuration interface used to configure the agent. After the user uploads the corresponding historical file, the corresponding content understanding model can be invoked according to the preset parsing dimensions. For example, the preset parsing dimensions can include event topic dimensions, person dimensions, etc. Taking chat logs as an example, after uploading the historical file, a corresponding prompt can be constructed based on the historical file, and the preset content understanding model can be used to understand it to determine the various sub-files corresponding to the historical file under the parsing dimensions. For example, under the event topic dimension, sub-file F1 can be divided into chat logs discussing food, and sub-file F2 can be divided into chat logs discussing movies, etc. The topic under the event topic dimension can be determined by parsing the historical file; that is, the content understanding model groups chat logs discussing the same topic together and determines its corresponding event topic to add a dimension identifier to the sub-file. For example, the dimension identifier for sub-file F1 could be the topic "food". Similarly, under the person dimension, chat logs between the target user and different people can be divided into different sub-files.

[0093] Correspondingly, the file information of a sub-file can include the dimension identifier corresponding to that sub-file, thereby allowing the user to clearly understand the summary of the content of the sub-file through the display of the file information, so that the user can make a selection. The file information of the multiple sub-files can be displayed separately in the display interface, allowing the user to make a selection based on the file information. For example, the display interface can show a selection box corresponding to each file information, and the sub-files under the user-selected file information can be identified as the user's selected sub-files. For instance, the user can select file information by checking the selection box. If the user selects the selection box corresponding to the theme "Food", then the sub-files under the dimension of "Food" can be identified as the user's selected sub-files, thereby determining the historical memory data.

[0094] As another example, if the interactive interface is a dialogue interface with the intelligent agent, after the user uploads a historical file through the dialog box, the historical file can be understood based on the method described above to obtain multiple sub-files. Then, the file information of these multiple sub-files can be displayed to the user as an output dialog, providing options such as, "This file can be divided into the following sub-files based on the event theme dimension: Z11: Theme - Food, Z12: Theme - Movie, ..., and divided into the following sub-files based on the person dimension: Z21: Person 1, Z22: Person 2, ..., please reply with the corresponding number to select the corresponding sub-file." The user can then further select the file information by entering the corresponding number in the dialog box. Based on the user's selected file information, the sub-files under the selected file information can be determined as the user-selected sub-files, thus further determining the historical memory data. For example, if the user enters the number Z11 in the dialog box, the sub-files under the theme - Food dimension can be determined as the user-selected sub-files, thereby determining the historical memory data.

[0095] Therefore, the above technical solution can automatically understand the content of data in historical files, thereby automatically splitting the historical files into multiple sub-files so that users can upload the corresponding data from the sub-files, realizing the uploading of partial data in the file. In this process, users do not need to manually split the file, saving user operations and simplifying the data uploading process in the file.

[0096] In some possible embodiments, the method further includes:

[0097] Before obtaining the target agent corresponding to the target user based on the historical files, the historical files are converted to the target format; the target agent corresponding to the target user is obtained based on the historical files in the target format.

[0098] Accordingly, in order to improve the efficiency of data upload and unify standards, the target format of file upload can be preset. As an example, the target format can be set separately for different data formats in historical files. For example, the target format for text data can be set to txt format, and the target format for audio data can be set to mp3 format. The setting can be based on the actual application scenario, and this disclosure does not limit it.

[0099] In this embodiment, when uploading historical files, taking text data as an example, if the format of the historical file is doc, it can be converted to txt, and then subsequent processing can be performed based on the txt format historical file.

[0100] Therefore, by using the above technical solution, the unified management and uploading of historical files can be achieved through the conversion of their formats, which facilitates the subsequent processing of historical files and improves data processing efficiency.

[0101] In one possible embodiment, an exemplary implementation of obtaining the target agent corresponding to the target user based on the historical memory data is as follows, and this step may include:

[0102] The agent is trained based on the training data corresponding to the historical memory data to obtain the target agent corresponding to the target user.

[0103] The training data corresponding to the historical memory data includes the historical memory data, or the training data corresponding to the historical memory data includes the historical memory data and generated data. The generated data is obtained by performing target processing on the historical memory data. The target processing includes at least one of the following: synonym replacement processing, word order adjustment processing, and semantic enhancement processing.

[0104] In training the agent, for example, the obtained historical memory data can be directly used as training data to ensure consistency between the target agent's performance characteristics and the target user's characteristics. As another example, to further improve the diversity of training data, new training data can be generated based on the historical memory data to enrich the agent's training dataset. For instance, semantic enhancement of the historical memory data can be achieved through target processing, such as synonym replacement. This processing involves randomly selecting words from sentences in the historical memory data and replacing them with synonyms to generate data. Word order adjustment processing involves randomly selecting sentences from the historical memory data and changing the order of words or phrases within those sentences to generate data. Semantic enhancement processing involves randomly selecting sentences from the historical memory data, querying relevant background information based on the sentence's semantics, and adding this background information to enhance the sentence's semantics to generate data. This process can involve semantic analysis of the sentence to query background information based on its semantics; this query can be performed using semantic search methods within the field and is not limited here. It should be noted that when generating data based on target processing, it is necessary to ensure that the meaning or label of the historical memory data and the generated data obtained based on the historical memory data are consistent. In this way, new training data can be obtained by changing the features of the historical memory data, thereby achieving effective expansion of the training dataset.

[0105] Therefore, through the above technical solutions, new generated data can be obtained based on existing historical memory data, thereby increasing the diversity and richness of the training dataset for the intelligent agent. Through data augmentation and fusion strategies, the efficiency of data use can be improved, thereby enhancing the accuracy and reliability of the model. To a certain extent, this can improve the consistency between the performance of the target intelligent agent and the target user, and enhance the user experience when the user interacts with the target intelligent agent.

[0106] In some possible embodiments, the target configuration interface also includes a language configuration item, as shown in A1 of Figure 2, which can be used by the user to select a language.

[0107] Accordingly, the method further includes:

[0108] In response to receiving the language configuration option selection operation, a language selection interface is displayed, which carries multiple candidate languages. The agent's speech output can be achieved through a text-to-speech (TTS) model. Candidate languages ​​can be any language supported by the speech synthesis model, such as Chinese and English. To further adapt to various application scenarios, the candidate languages ​​can also include multiple dialects.

[0109] As an example, this speech synthesis model can be determined in the following way:

[0110] The training audio data is obtained through pre-labeling and filtering. This pre-selected audio data can be high-quality audio data, such as audio data from public speeches or interviews, thereby improving the quality of the training audio data. The filtering can be based on the actual application scenario, and this disclosure does not limit it in this regard.

[0111] Next, the training audio data is processed to obtain processed audio data. This processing can include speed adjustment, pitch adjustment, etc., thus obtaining new training data without re-annotation, increasing the diversity of the training data and improving the robustness of the speech synthesis model. Furthermore, the speech synthesis model can be trained based on this training audio data, the processed audio data, and their corresponding text to obtain the speech synthesis model. In this speech synthesis model, the linguistic parameters can be optimized by incorporating analysis of different dialects in this field to improve the adaptability of the speech synthesis model to different dialect features.

[0112] In response to receiving a selection operation for a candidate language, the candidate language corresponding to the selection operation is determined as the target language when the target agent outputs data.

[0113] Users can determine the language that the target agent needs to output from the candidate languages ​​based on their own needs. For example, if the target agent needs to output Mandarin, the user can select Mandarin from the candidate languages ​​and confirm it. This will determine Mandarin as the target language when the target agent outputs data. That is, when the user interacts with the target agent, the target agent will output a reply in Mandarin, such as text or voice in Mandarin.

[0114] Therefore, through the above technical solution, the language used by the target intelligent agent when interacting with the user can be pre-configured, so that when the user interacts with the target intelligent agent, the target intelligent agent can output the language type required by the user, so that the generated target intelligent agent can meet the needs of a wider range of users.

[0115] In some possible embodiments, the historical file contains audio data, and the target agent outputs speech corresponding to the target language based on a speech synthesis model;

[0116] The method may further include:

[0117] The parameters of the speech synthesis model are updated based on the audio data.

[0118] As an example, further, to improve the consistency between the target agent's output speech and the target user's speech, the parameters of the speech synthesis model can be fine-tuned and updated based on the uploaded audio data. This can be achieved by enabling the speech synthesis model to generate speech that matches the voice characteristics of the target user. For instance, a diffusion model can be used, employing the audio data as a sample speech prompt to adjust the parameters of the language synthesis model, so that when the user interacts with the target agent subsequently, the language synthesis model outputs speech with a rhythm and style consistent with the target user.

[0119] Therefore, through the above technical solution, during the training process of the target intelligent agent, the parameters of the speech synthesis model can be adjusted simultaneously with the audio data uploaded by the user, so that the speech output based on the speech synthesis model is consistent with the characteristics and style of the target user's speech. This improves the simulation and reproduction of the target user's personality and behavioral characteristics by the target intelligent agent, making the user's interaction with the target intelligent agent more like a real interaction between the user and the target user. It also improves the anthropomorphism and personalization of the target intelligent agent's interaction, and to a certain extent satisfies the user's emotional needs for interaction.

[0120] In one possible embodiment, the method may further include:

[0121] The system receives descriptive information about the target intelligent agent. When configuring the target intelligent agent, the user can input descriptive information, which can be entered through a text input box, as shown at A3 in Figure 2. This descriptive information can be information characterizing the personality traits and / or behavioral characteristics of the target intelligent agent, and the user can describe and input it based on their needs.

[0122] As an example, the target configuration interface can also include configuration items such as name and image, as shown in A4 of Figure 2. Users can configure these configuration items to generate the attributes corresponding to the target agent and store them in association with the target agent. For example, users can select to interact with the corresponding target agent based on the name attribute.

[0123] Accordingly, obtaining the target agent corresponding to the target user based on the historical memory data includes:

[0124] The agent is trained based on the historical memory data and the descriptive information to obtain the target agent corresponding to the target user.

[0125] Accordingly, the historical memory data and descriptive information can be used as training data to train an artificial intelligence model based on machine learning and deep learning technologies. The resulting artificial intelligence model can then be used as the target intelligent agent.

[0126] Therefore, by using the above technical solution, the intelligent agent can be trained by combining the user's feature description information of the target intelligent agent and the target user's historical memory data. This will enable the trained target intelligent agent to not only have similar response characteristics and behavioral characteristics to the target user, but also to better match the user's interaction needs for the target intelligent agent, thereby further improving the user experience.

[0127] Based on the same inventive concept, this disclosure also provides an interaction method, as shown in Figure 3, which may include:

[0128] In step 31, user input information is received. This input information is provided by the user during interaction, and the input type can include text, voice, video, etc.

[0129] In step 32, based on the target agent, the response information corresponding to the input information is determined, wherein the target agent is determined based on any of the agent generation methods described above.

[0130] The input information can be input into the target intelligent agent, which will then analyze and process the input information to obtain the corresponding response information, which can be text information.

[0131] In step 33, a response message is output based on the input type of the input information.

[0132] As an example, the input type can be used as the output type for the reply message, thus outputting that reply message. For instance, if the input message is "Have you recommended any TV series lately?", and the determined reply message is "I've been watching XXX recently, and I think it's pretty good; the plot is quite fast-paced," then if the input type is text dialogue, the reply message can be output as text dialogue, meaning it can be output as text in the chat interface. If the input type is voice dialogue, the reply message can be output as voice dialogue, meaning the reply message can be used to generate voice, and that voice can be output as voice in the chat interface.

[0133] As another example, the current output type can be determined based on the frequency of the target agent's use of different output types. For instance, the current output type can be determined by sampling from different output types based on the frequency of use. Alternatively, the output type most frequently used between the target agent and the current user can be used as the output type corresponding to the response message.

[0134] Therefore, through the above technical solution, interaction with the user can be achieved based on the target intelligent agent. The target intelligent agent is determined based on the user's needs and through the target user's historical files, thereby ensuring the validity and reliability of the response information determined by the target intelligent agent, improving the user's satisfaction with the interaction, and at the same time, enhancing the user's interest in interaction to a certain extent and providing emotional support for the user.

[0135] Figure 4 shows a UML diagram corresponding to the interaction method provided in the embodiments of this disclosure. UML diagrams, or Unified Modeling Language diagrams, are a language used for visual modeling of software systems. For example, the class implementing this interaction method can include a MessageViewModel class, which can contain related attributes and methods such as chat room data, message database, WebSocket source, and HTTP source. The HTTP source can be used to receive data transmitted via HTTP (Hypertext Transfer Protocol), and WebSocket is a protocol for full-duplex communication over a single TCP connection; the WebSocket source can be used to implement bidirectional communication. As shown in Figure 4, the message display class can contain the following methods:

[0136] The `prepare` method can be used to set up the WebSocket listener and retrieve previous chat history from the message database. The `fetchVoice` method retrieves audio content based on the text corresponding to the message list index, used for voice playback. The `fetchTalkId` method generates a video stream based on the audio content and the user's image, returning the conversation ID. The `fetchTalkUrl` method retrieves the conversation URL based on the conversation ID, allowing the retrieval of the video stream for display. The `addMessage` method adds new messages to the message list and saves them to the message database. The `fetchRoomMessages` method retrieves the chat room's message history from the message database. The `saveRoomMessages` method saves messages to the message database. The `sendMessage` method is used for users to send messages, and the `close` method ends the conversation. The implementation of these methods can be based on corresponding processing logic in the art, and this disclosure does not limit their implementation.

[0137] Accordingly, classes such as chat room data, message database, WebSocket source, HTTP source, and message data can further include corresponding fields and methods. For example, the chat room data class can further include fields such as chat room ID and chat room title. The message data class can include fields such as message identifier, text content, audio content, dialogue ID, and dialogue URL. These can be set based on the actual application scenario, and this disclosure does not impose any limitations on them.

[0138] As an example, Figure 5 shows a flowchart of text or voice interaction in a chat page provided by an embodiment of this disclosure. For example, when a user interacts with a target agent, the terminal performs an initialization operation (calling the prepare method), retrieves the chat history of the current chat room from the message database (calling the fetchRoomMessages method), then retrieves the message sent by the user (calling the sendMessage method), determines the corresponding reply message based on the target agent, and then determines whether to generate audio content (calling the fetchVoice method). If audio content is generated, the voice is played to reply to the user; if no audio content is generated, a text message is output to reply to the user. After replying to the user, the reply message can be added to the message list (calling the addMessage method) and saved to the message database (calling the saveRoomMessages method). Simultaneously, the terminal can continue to retrieve information sent by the user and repeat the above process to engage in dialogue with the user. When the dialogue is complete, the dialogue process can be ended, such as by calling the close method described above.

[0139] As an example, Figure 6 shows a flowchart of video interaction in a chat page provided by an embodiment of this disclosure. For example, when a user interacts with a target agent, the terminal performs an initialization operation (calling the prepare method), retrieves the chat history of the current chat room from the message database (calling the sendMessage method), then retrieves the message sent by the user (calling the sendMessage method), determines the reply message corresponding to the message based on the target agent, then determines whether to generate audio content (calling the fetchVoice method). If no audio content is generated, a text message is output to reply to the user. If audio content is generated, a video stream is further generated based on the audio content and the image of the target agent, and a dialogue ID is returned (calling the fetchTalkId method). The URL of the dialogue is obtained based on the dialogue ID, and the video stream is fetched and displayed based on the URL (calling the fetchTalkUrl method). The video is played to reply to the user. After replying to the user, the reply message can be added to the message list (calling the addMessage method) and the message can be saved to the message database (calling the saveRoomMessages method). At the same time, the terminal can continue to retrieve information sent by the user and repeat the above process to have a dialogue with the user. The conversation can be ended when it is finished.

[0140] Based on the same inventive concept, this disclosure also provides an intelligent agent generation device, as shown in FIG7, wherein the device 10 includes:

[0141] The first display module 101 is used to display the interactive interface corresponding to the intelligent agent;

[0142] The acquisition module 102 is used to obtain the historical file of the target user corresponding to the intelligent agent based on the interactive interface. The historical file contains the operation records of the target user within the historical time period.

[0143] The processing module 103 is used to obtain the target intelligent agent corresponding to the target user based on the historical files.

[0144] Optionally, the interactive interface is a target configuration interface for configuring the intelligent agent;

[0145] The acquisition module includes:

[0146] The first display submodule is used to display the historical data upload interface in response to the selection operation of the historical data configuration item in the target configuration interface;

[0147] The first acquisition submodule is used to use the files received based on the data upload interface as the historical files.

[0148] Optionally, the interactive interface is an interface for engaging in dialogue with the intelligent agent;

[0149] The acquisition module includes:

[0150] The second acquisition submodule is used to receive the target user's historical files from the interactive interface during the dialogue process of the agent.

[0151] Optionally, the processing module includes:

[0152] The first determining submodule is used to determine the historical memory data corresponding to the intelligent agent based on the historical file, wherein the historical memory data is at least a part of the data in the historical file;

[0153] The first processing submodule is used to obtain the target intelligent agent corresponding to the target user based on the historical memory data.

[0154] Optionally, the first determining submodule includes:

[0155] The second display submodule is used to display the processing interface of the historical files;

[0156] The second processing submodule is used to respond to receiving the user's editing operation on the historical file in the processing interface, and to use the data in the file obtained after editing the historical file as the historical memory data.

[0157] And / or in response to receiving a user's selection operation on the time selection control in the processing interface, determine the target time period and determine the data within the target time period in the history file as the historical memory data.

[0158] Optionally, the first determining submodule includes:

[0159] The parsing submodule is used to parse the historical file to obtain multiple sub-files corresponding to the historical file;

[0160] The third display submodule is used to display the file information of the multiple sub-files;

[0161] The second determining submodule responds to determining the subfile selected by the user and uses the data of the selected subfile as the historical memory data.

[0162] Optionally, the first processing submodule includes:

[0163] The training submodule is used to train the agent based on the training data corresponding to the historical memory data to obtain the target agent corresponding to the target user.

[0164] The training data corresponding to the historical memory data includes the historical memory data, or the training data corresponding to the historical memory data includes the historical memory data and generated data. The generated data is obtained by performing target processing on the historical memory data. The target processing includes at least one of the following: synonym replacement processing, word order adjustment processing, and semantic enhancement processing.

[0165] Optionally, the device further includes:

[0166] The first receiving module is used to receive description information of the target intelligent agent;

[0167] The first processing submodule is further configured to:

[0168] The agent is trained based on the historical memory data and the descriptive information to obtain the target agent corresponding to the target user.

[0169] Optionally, the target configuration interface may also include a language configuration item;

[0170] The device further includes:

[0171] The second display module is used to display a language selection interface in response to receiving the language configuration item selection operation, wherein the language selection interface carries multiple candidate languages;

[0172] The first determining module is configured to, in response to receiving a selection operation for a candidate language, determine the candidate language corresponding to the selection operation as the target language corresponding to the output data of the target agent.

[0173] Optionally, the historical files contain audio data, and the target agent outputs speech based on a speech synthesis model;

[0174] The device further includes:

[0175] An update module is used to update the parameters of the speech synthesis model based on the audio data.

[0176] Optionally, the historical files may contain at least one of audio data, text data, video data, and published content.

[0177] Based on the same inventive concept, this disclosure also provides an interactive device, as shown in FIG8, wherein the device 20 includes:

[0178] The second receiving module 201 is used to receive user input information;

[0179] The second determining module 202 is used to determine the response information corresponding to the input information based on the target intelligent agent, wherein the target intelligent agent is determined based on any of the intelligent agent generation methods described above;

[0180] The output module 203 is used to output the response information based on the input type of the input information.

[0181] Referring now to FIG9, a schematic diagram of the structure of an electronic device (e.g., a terminal device or a server) 600 suitable for implementing embodiments of the present disclosure is shown. The terminal device in the embodiments of the present disclosure may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. The electronic device shown in FIG9 is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of the present disclosure.

[0182] As shown in Figure 9, the electronic device 600 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage device 608 into a random access memory (RAM) 603. The RAM 603 also stores various programs and data required for the operation of the electronic device 600. The processing unit 601, ROM 602, and RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0183] Typically, the following devices can be connected to I / O interface 605: input devices 606 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 607 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 608 including, for example, magnetic tapes, hard disks, etc.; and communication devices 609. Communication device 609 allows electronic device 600 to communicate wirelessly or wiredly with other devices to exchange data. Although FIG9 shows electronic device 600 with various devices, it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed alternatively.

[0184] In particular, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 609, or installed from a storage device 608, or installed from a ROM 602. When the computer program is executed by the processing device 601, it performs the functions defined in the methods of embodiments of this disclosure.

[0185] It should be noted that the computer-readable medium described in this disclosure can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this disclosure, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, apparatus, or device. In this disclosure, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0186] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol), and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and end-to-end networks (e.g., ad hoc end-to-end networks), as well as any currently known or future-developed networks.

[0187] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device.

[0188] The aforementioned computer-readable medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to: display an interactive interface corresponding to the intelligent agent; obtain a historical file of the target user corresponding to the intelligent agent based on the interactive interface, the historical file containing the operation records of the target user within a historical period; and obtain a target intelligent agent corresponding to the target user based on the historical file.

[0189] Alternatively, the aforementioned computer-readable medium carries one or more programs that, when executed by the electronic device, cause the electronic device to: receive user input information; determine response information corresponding to the input information based on a target intelligent agent, wherein the target intelligent agent is determined based on the intelligent agent generation method described in any of the above embodiments; and output the response information based on the input type of the input information.

[0190] Computer program code for performing the operations of this disclosure can be written in one or more programming languages ​​or a combination thereof, including but not limited to object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0191] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0192] The modules described in the embodiments of this disclosure can be implemented in software or in hardware. The names of the modules are not necessarily limiting in certain circumstances; for example, the first display module can also be described as "a module that displays the interactive interface corresponding to the intelligent agent".

[0193] The functions described above in this document can be performed, at least in part, by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: Field Programmable Gate Arrays (FPGAs), Application-Specific Integrated Circuits (ASICs), Application Standard Products (ASSPs), System-on-Chip (SoCs), Complex Programmable Logic Devices (CPLDs), and so on.

[0194] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0195] According to one or more embodiments of this disclosure, Example 1 provides a method for generating an intelligent agent, wherein the method includes:

[0196] Display the interactive interface corresponding to the intelligent agent;

[0197] Based on the interactive interface, the historical file of the target user corresponding to the intelligent agent is obtained, and the historical file contains the operation record of the target user within the historical time period;

[0198] Based on the historical files, the target intelligent agent corresponding to the target user is obtained.

[0199] According to one or more embodiments of this disclosure, Example 2 provides the method of Example 1, wherein the interactive interface is a target configuration interface for configuring the agent;

[0200] The step of obtaining the historical files of the target user corresponding to the intelligent agent based on the interactive interface includes:

[0201] In response to the selection operation of the historical data configuration item in the target configuration interface, the historical data upload interface is displayed;

[0202] The files received based on the data upload interface are used as the historical files.

[0203] According to one or more embodiments of this disclosure, Example 3 provides the method of Example 1, wherein the interactive interface is an interface for engaging in dialogue with the intelligent agent;

[0204] The step of obtaining the historical files of the target user corresponding to the intelligent agent based on the interactive interface includes:

[0205] During the dialogue process of the intelligent agent, the target user's historical files are received from the interactive interface.

[0206] According to one or more embodiments of this disclosure, Example 4 provides the method of Example 1, wherein obtaining the target agent corresponding to the target user based on the historical file includes:

[0207] Based on the historical files, the historical memory data corresponding to the intelligent agent is determined, wherein the historical memory data is at least a portion of the data in the historical files;

[0208] Based on the historical memory data, the target intelligent agent corresponding to the target user is obtained.

[0209] According to one or more embodiments of this disclosure, Example 5 provides the method of Example 4, wherein determining the historical memory data corresponding to the agent based on the historical file includes:

[0210] Display the interface for processing the historical files;

[0211] In response to receiving an editing operation from the user on the processing interface for the historical file, the data in the file obtained after editing the historical file is used as the historical memory data;

[0212] And / or in response to receiving a user's selection operation on the time selection control in the processing interface, determine the target time period and determine the data within the target time period in the history file as the historical memory data.

[0213] According to one or more embodiments of this disclosure, Example 6 provides the method of Example 4, wherein determining the historical memory data corresponding to the agent based on the historical file includes:

[0214] The historical file is parsed to obtain multiple sub-files corresponding to the historical file;

[0215] Display the file information of the multiple sub-files;

[0216] In response to determining the sub-file selected by the user, the data of the selected sub-file is used as the historical memory data.

[0217] According to one or more embodiments of this disclosure, Example 7 provides the method of Example 4, wherein obtaining the target agent corresponding to the target user based on the historical memory data includes:

[0218] The agent is trained based on the training data corresponding to the historical memory data to obtain the target agent corresponding to the target user;

[0219] The training data corresponding to the historical memory data includes the historical memory data, or the training data corresponding to the historical memory data includes the historical memory data and generated data. The generated data is obtained by performing target processing on the historical memory data. The target processing includes at least one of the following: synonym replacement processing, word order adjustment processing, and semantic enhancement processing.

[0220] According to one or more embodiments of this disclosure, Example 8 provides the method of Example 4, wherein the method further includes:

[0221] Receive description information of the target intelligent agent;

[0222] The step of obtaining the target intelligent agent corresponding to the target user based on the historical memory data includes:

[0223] The agent is trained based on the historical memory data and the descriptive information to obtain the target agent corresponding to the target user.

[0224] According to one or more embodiments of this disclosure, Example 9 provides the method of Example 2, wherein the target configuration interface further includes a language configuration item;

[0225] The method further includes:

[0226] In response to receiving the language configuration item selection operation, a language selection interface is displayed, which carries multiple candidate languages;

[0227] In response to receiving a selection operation for a candidate language, the candidate language corresponding to the selection operation is determined as the target language when the target agent outputs data.

[0228] According to one or more embodiments of this disclosure, Example 10 provides the method of Example 1, wherein the history file contains audio data, and the target agent outputs speech based on a speech synthesis model;

[0229] The method further includes:

[0230] The parameters of the speech synthesis model are updated based on the audio data.

[0231] According to one or more embodiments of this disclosure, Example 11 provides the method of Example 1, wherein the history file contains at least one of audio data, text data, video data, and published content.

[0232] According to one or more embodiments of this disclosure, Example 12 provides an interaction method, the method comprising:

[0233] Receive user input information;

[0234] Based on the target agent, determine the response information corresponding to the input information, wherein the target agent is determined based on any of the agent generation methods described in Examples 1-11;

[0235] Based on the input type of the input information, the response information is output.

[0236] According to one or more embodiments of this disclosure, Example 13 provides an apparatus for generating an intelligent agent, the apparatus comprising:

[0237] The first display module is used to display the interactive interface corresponding to the intelligent agent;

[0238] The acquisition module is used to obtain the historical file of the target user corresponding to the intelligent agent based on the interactive interface. The historical file contains the operation records of the target user within the historical time period.

[0239] The processing module is used to obtain the target intelligent agent corresponding to the target user based on the historical files.

[0240] According to one or more embodiments of this disclosure, Example 14 provides an interactive device, the device comprising:

[0241] The second receiving module is used to receive user input information;

[0242] The second determining module is used to determine the response information corresponding to the input information based on the target intelligent agent, wherein the target intelligent agent is determined based on the intelligent agent generation method according to any one of claims 1-11;

[0243] The output module is used to output the response information based on the input type of the input information.

[0244] According to one or more embodiments of the present disclosure, Example 15 provides a computer-readable medium having a computer program stored thereon that, when executed by a processing device, implements the steps of the method described in any one of Examples 1-12.

[0245] According to one or more embodiments of this disclosure, Example 16 provides an electronic device comprising:

[0246] A storage device on which computer programs are stored;

[0247] A processing device for executing the computer program in the storage device to implement the steps of any one of the methods in Examples 1-12.

[0248] According to one or more embodiments of the present disclosure, Example 17 provides a computer program product including a computer program that, when executed by a processor, implements the steps of the method described in any one of Examples 1-12.

[0249] The above description is merely a preferred embodiment of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features disclosed in this disclosure that have similar functions.

[0250] Furthermore, while the operations are described in a specific order, this should not be construed as requiring these operations to be performed in the specific order shown or in a sequential order. In certain environments, multitasking and parallel processing may be advantageous. Similarly, while several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of this disclosure. Certain features described in the context of individual embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented individually or in any suitable sub-combination in multiple embodiments.

[0251] Although the subject matter has been described using language specific to structural features and / or methodological logic, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Rather, the specific features and actions described above are merely illustrative forms of implementing the claims. Regarding the apparatus in the above embodiments, the specific manner in which the various modules perform their operations has been described in detail in the embodiments relating to the method, and will not be elaborated upon here.

Claims

1. A method for generating an agent, comprising: displaying an interaction interface corresponding to an agent; obtaining a history file of a target user corresponding to the agent based on the interaction interface, the history file containing operation records of the target user in a historical period; obtaining a target agent corresponding to the target user based on the history file.

2. The method of claim 1, wherein, The interaction interface is a target configuration interface for configuring the agent. The obtaining of the history file of the target user corresponding to the agent based on the interaction interface comprises: in response to a selection operation on a historical data configuration item in the target configuration interface, displaying a historical data upload interface; receiving a file based on the data upload interface as the history file.

3. The method of claim 1, wherein, The interaction interface is an interface for a conversation with the agent. The obtaining of the history file of the target user corresponding to the agent based on the interaction interface comprises: receiving a history file of the target user input by a user from the interaction interface during a conversation process of the agent.

4. The method according to any one of claims 1 to 3, wherein, The obtaining of the target agent corresponding to the target user based on the history file comprises: determining historical memory data corresponding to the agent based on the history file, the historical memory data being at least part of data in the history file; obtaining the target agent corresponding to the target user based on the historical memory data.

5. The method of claim 4, wherein, The determining of the historical memory data corresponding to the agent based on the history file comprises: displaying a processing interface of the history file; in response to receiving an editing operation of the history file by a user in the processing interface, determining data in a file obtained after editing the history file as the historical memory data; and / or in response to receiving a selection operation of a time selection control by a user in the processing interface, determining a target time period, and determining data in the target time period in the history file as the historical memory data.

6. The method of claim 4, wherein, The determining of the historical memory data corresponding to the agent based on the history file comprises: parsing the history file to obtain a plurality of sub-files corresponding to the history file; displaying file information of the plurality of sub-files; in response to determining that a sub-file selected by a user, determining data of the selected sub-file as the historical memory data.

7. The method of claim 4, wherein, The obtaining of the target agent corresponding to the target user based on the historical memory data comprises: training the agent based on training data corresponding to the historical memory data to obtain the target agent corresponding to the target user; the training data corresponding to the historical memory data contains the historical memory data, or the training data corresponding to the historical memory data contains the historical memory data and generated data, the generated data being obtained by target processing of the historical memory data, the target processing containing at least one of the following: synonym replacement processing, syntax adjustment processing, semantic enhancement processing.

8. The method of any one of claims 4-7, further comprising: receiving description information of the target agent. The target intelligent agent corresponding to the target user is obtained based on the historical memory data. The target intelligent agent corresponding to the target user is obtained based on the historical memory data and the description information.

9. The method of claim 2, wherein, The target configuration interface further includes a language configuration item; The method further includes: In response to receiving a selection operation of the language configuration item, a language selection interface is displayed, and the language selection interface carries a plurality of candidate languages; In response to receiving a selection operation of a candidate language, the candidate language corresponding to the selection operation is determined as a target language corresponding to output data of the target intelligent agent.

10. The method of any one of claims 1-9, wherein, The historical file includes audio data, and the target intelligent agent outputs speech based on a speech synthesis model; The method further includes: Parameters of the speech synthesis model are updated based on the audio data.

11. The method according to any one of claims 1-9, wherein, The historical file includes at least one of audio data, text data, video data, and published content.

12. An interaction method, comprising: receiving input information of a user; determining reply information corresponding to the input information based on a target intelligent agent, wherein the target intelligent agent is determined based on the generation method of the intelligent agent in any one of claims 1-11; outputting the reply information based on an input type of the input information.

13. An intelligent agent generation apparatus, comprising: a first display module configured to display an interaction interface corresponding to an intelligent agent; an acquisition module configured to acquire a historical file of a target user corresponding to the intelligent agent based on the interaction interface, the historical file including operation records of the target user in a historical period; a processing module configured to obtain a target intelligent agent corresponding to the target user based on the historical file.

14. An interaction apparatus, comprising: a second receiving module configured to receive input information of a user; a second determination module configured to determine reply information corresponding to the input information based on a target intelligent agent, wherein the target intelligent agent is determined based on the generation method of the intelligent agent in any one of claims 1-11; an output module configured to output the reply information based on an input type of the input information.

15. A computer readable medium storing a computer program, wherein, The computer program is executed by the processing apparatus to implement the method in any one of claims 1-12.

16. An electronic device, comprising: a storage device storing a computer program; a processing apparatus configured to execute the computer program in the storage device to implement the generation method of the intelligent agent in any one of claims 1-11 or the interaction method in claim 12.

17. A computer program product comprising a computer program, wherein, The computer program is executed by the processor to implement the generation method of the intelligent agent in any one of claims 1-11 or the interaction method in claim 12.

Images