Brand operation method and device, electronic equipment, storage medium and computer program product

Abstract

The embodiment of the invention provides a brand operation method and device, electronic equipment, a storage medium and a computer program product. In the embodiment of the invention, the intelligent agent demand information of the brand operator is obtained, and the intelligent agent matched with the demand is selected from a plurality of basic intelligent agents, so that the selection of the intelligent agent is ensured to have pertinence and applicability; and according to the intelligent agent demand information of the brand operator, carrying out detailed configuration on the selected basic version intelligent agent to obtain a special intelligent agent which can meet the specific demand of the brand operator. The special intelligent agent can provide highly personalized services, meet specific requirements of brand operators, improve service quality and customer satisfaction, improve efficiency and quality of brand operation, significantly improve the overall level of brand operation, and enhance market competitiveness of brands.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a brand operation method, device, electronic device, storage medium and computer program product. Background Art

[0002] With the rapid development of artificial intelligence technology, the complexity and variability of brand operations require a higher level of intelligence. Traditional brand operation methods can no longer meet the dynamic needs of the modern market, and a new intelligent solution is urgently needed to improve operational efficiency and effectiveness. Summary of the invention

[0003] Multiple aspects of the present application provide a brand operation method, device, electronic device, storage medium and computer program product to improve brand operation efficiency and effectiveness.

[0004] An embodiment of the present application provides a brand operation method, including: obtaining smart agent demand information of a brand operator, wherein the smart agent demand information includes at least one smart agent type required by the brand operator, and working parameters corresponding to each smart agent type; selecting at least one first basic version smart agent that matches at least one smart agent type required by the brand operator from multiple basic version smart agents; configuring at least one first basic version smart agent according to the working parameters corresponding to at least one smart agent type required by the brand operator to obtain at least one dedicated smart agent; and calling at least one dedicated smart agent to provide brand operation services for the brand operator.

[0005] An embodiment of the present application also provides a brand operation device, including: an acquisition module, used to obtain smart agent demand information of a brand operator, wherein the smart agent demand information includes at least one smart agent type required by the brand operator, and working parameters corresponding to each smart agent type; a selection module, used to select at least one first basic version smart agent that matches at least one smart agent type required by the brand operator from multiple basic version smart agents; a configuration module, used to configure at least one first basic version smart agent according to the working parameters corresponding to at least one smart agent type required by the brand operator, to obtain at least one dedicated smart agent; and a calling module, used to call at least one dedicated smart agent to provide brand operation services for the brand operator.

[0006] An embodiment of the present application also provides an electronic device, including: a memory and a processor; the memory is used to store a computer program; the processor is coupled to the memory, and is used to execute the computer program to execute the steps in the brand operation method.

[0007] An embodiment of the present application also provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor is enabled to implement the steps in the brand operation method.

[0008] An embodiment of the present application also provides a computer program product, including a computer program / instruction, which, when executed by a processor, enables the processor to implement the steps in the brand operation method.

[0009] In the technical solution provided in the embodiment of the present application, the smart agent demand information of the brand operator is obtained, and the smart agent matching the demand is selected from multiple basic version smart agents, ensuring that the selection of the smart agent is targeted and applicable; according to the smart agent demand information of the brand operator, the selected basic version smart agent is configured in detail to obtain a dedicated smart agent that can meet the specific needs of the brand operator. Dedicated smart agents can provide highly personalized services, meet the specific needs of brand operators, improve service quality and customer satisfaction, improve the efficiency and quality of brand operations, significantly improve the overall level of brand operations, and enhance the market competitiveness of brands. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0011] Figure 1 A flow chart of a brand operation method provided in an embodiment of the present application;

[0012] Figure 2 A flow chart of a brand operation method provided in an embodiment of the present application;

[0013] Figure 3a This is a schematic diagram of the structure of an exemplary brand operation full-link intelligent body base;

[0014] Figure 3b A schematic diagram of the structure of a brand operation device provided in an embodiment of the present application;

[0015] Figure 4 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.

[0017] In the embodiments of the present application, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the access relationship of associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist at the same time, and B exists alone, where A and B may be singular or plural. In the text description of the present application, the character " / " generally indicates that the previous and next associated objects are in an "or" relationship. In addition, in the embodiments of the present application, "first", "second", "third", etc. are only used to distinguish the contents of different objects and have no other special meanings.

[0018] The following specific embodiments are used to describe in detail the technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The following is a detailed description of the technical solutions provided by each embodiment of the present application in conjunction with the accompanying drawings.

[0019] Figure 1 A flowchart of a brand operation method provided in an embodiment of the present application. Figure 1 , the method may include the following steps:

[0020] 101. Obtain smart agent demand information of a brand operator, wherein the smart agent demand information includes at least one smart agent type required by the brand operator and working parameters corresponding to each smart agent type.

[0021] In actual applications, the brand operator's intelligent agent demand information can be obtained through various methods, for example, by inviting the brand operator to fill out an online questionnaire, directly communicating with the brand operator through telephone, email, instant messaging tools, or through API (Application Programming Interface) interface automated acquisition to obtain the brand operator's intelligent agent demand information, and there is no restriction on this.

[0022] In actual applications, the brand operator's intelligent agent demand information reflects the relevant information of the intelligent agent (Agent) required by the brand operator. The intelligent agent can also be called an intelligent agent. An intelligent agent refers to an autonomous entity that can perceive the environment and influence the environment by performing actions. The intelligent agent can make decisions based on the information received and take actions to achieve the predetermined goals. In brand operations, intelligent agents can play a variety of roles, such as customer service representatives, legal assistants, market analysts, shopping guides, etc. Intelligent agents can improve operational efficiency and service quality in an automated and intelligent way.

[0023] In actual applications, the smart agent demand information of the brand operator includes, for example, but is not limited to: at least one smart agent type required by the brand operator and the working parameters corresponding to each smart agent type.

[0024] Optionally, the types of intelligent agents include, but are not limited to: intelligent customer service agent Agent, legal assistant Agent, networking assistant Agent and intelligent shopping guide Agent.

[0025] Among them, the intelligent customer service agent is an intelligent entity used to provide customer service, such as handling customer inquiries, answering common questions, order management and tracking, troubleshooting, transferring to manual customer service, etc.

[0026] Legal assistant agents are intelligent entities used to provide legal services, such as providing legal consulting services, contract review, case research, legal document generation, compliance checking, etc.

[0027] The Network Assistant Agent is an intelligent entity specially designed to help enterprises or individuals establish and maintain business connections and manage customer relationships. It can improve the efficiency and effectiveness of interpersonal network management through automation and intelligent means.

[0028] In this embodiment, the working parameters of the intelligent agent refer to the various settings and properties that need to be configured and adjusted when performing tasks, which directly affect the behavior and performance of the intelligent agent. For intelligent agents of different types and application scenarios, the working parameters may be different, and there is no limitation on this.

[0029] Optionally, the working parameters of the intelligent agent include, but are not limited to: working hours, performance parameters, retry mechanism, etc.

[0030] Among them, working hours refer to the working time period of the intelligent agent, such as 24 hours or 9:00-18:00 on weekdays.

[0031] Among them, performance parameters include, but are not limited to: processing speed, concurrency or resource usage, etc. Processing speed refers to the speed at which the agent processes tasks, such as how many pieces of data are processed per second. Concurrency refers to the ability of the agent to process multiple tasks at the same time. Resource usage refers to the computing resources occupied by the agent when it is running, such as CPU, memory, etc.

[0032] Among them, the retry mechanism refers to the retry strategy of the intelligent agent when encountering an error, such as the number of retries, interval time, etc. That is, the number of times and interval time that the intelligent agent retries after failure, such as a maximum of 3 retries, each with an interval of 5 seconds.

[0033] 102. Select at least one first basic version intelligent agent from among multiple basic version intelligent agents that matches at least one intelligent agent type required by a brand operator.

[0034] In the process of brand operation, the application of intelligent agents can significantly improve operational efficiency and service quality. In order to meet the diverse needs of different brand operators, multiple basic intelligent agents have been created in advance. These basic intelligent agents are universal and general and can adapt to a variety of business scenarios. The types of basic intelligent agents include, but are not limited to: basic intelligent customer service agent, basic legal assistant agent, basic networking assistant agent, and basic intelligent shopping guide agent.

[0035] According to at least one smart agent type required by the brand operator, the basic version of the smart agent required by the brand operator can be quickly selected from multiple basic version smart agents. Here, the basic version of the smart agent required by the brand operator is referred to as the first basic version of the smart agent.

[0036] 103. According to the working parameters corresponding to at least one intelligent agent type required by the brand operator, configure at least one first basic version intelligent agent to obtain at least one dedicated intelligent agent.

[0037] In actual applications, by making detailed configurations of the first basic version of the intelligent agent, it can be ensured that the generated dedicated intelligent agent can better adapt to the specific business scenarios of the brand operator, provide more personalized services, and improve the overall level of brand operation.

[0038] Taking the first basic version of the intelligent agent as an example, the basic version of the intelligent customer service agent is configured according to the business schedule of the brand operator, such as 24-hour service or 9:00-18:00 on weekdays; performance parameters such as processing speed, concurrency and resource usage are adjusted to optimize the performance of the intelligent agent to ensure that it can still run stably under high load conditions; set the retry strategy of the intelligent agent when encountering an error, such as the number of retries and the interval time, to ensure that the task can be completed smoothly.

[0039] 104. Call at least one dedicated intelligent agent to provide brand operation services to the brand operator.

[0040] In this embodiment, a dedicated intelligent agent is called to provide brand operation services for the brand operator. For example, the dedicated intelligent agent is an intelligent customer service agent, a legal assistant agent, a network assistant agent, and an intelligent shopping guide agent. The intelligent customer service agent is used to provide customer service during the brand operation process, for example, to provide 24-hour uninterrupted customer service, quickly handle customer inquiries and common problems, and improve customer satisfaction. The legal assistant agent is used to provide legal services during the brand operation process, for example, to provide efficient legal consulting services and contract reviews, reduce legal risks, and ensure compliance. The network assistant agent is used to establish and maintain business contacts and customer relationships during the brand operation process, promote business development, etc. The intelligent shopping guide agent is used to provide product shopping guide services, provide personalized shopping recommendations and shopping guidance, enhance user shopping experience, and increase sales.

[0041] The technical solution provided in the embodiment of the present application obtains the smart agent demand information of the brand operator, selects the smart agent that matches the demand from multiple basic version smart agents, and ensures that the selection of the smart agent is targeted and applicable; according to the smart agent demand information of the brand operator, the selected basic version smart agent is configured in detail to obtain a dedicated smart agent that can meet the specific needs of the brand operator. Dedicated smart agents can provide highly personalized services, meet the specific needs of brand operators, improve service quality and customer satisfaction, improve the efficiency and quality of brand operations, significantly improve the overall level of brand operations, and enhance the market competitiveness of brands.

[0042] In some optional embodiments, before obtaining the brand operator's intelligent agent demand information, it may also include: in response to a process orchestration operation on the process orchestration interface for any basic version of the intelligent agent to be constructed, calling the process orchestration service to orchestrate the work processing flow of the basic version of the intelligent agent; in response to a creation operation initiated on the intelligent agent creation interface, obtaining the configuration information of the basic version of the intelligent agent to be constructed, the configuration information including the identifier of the target open source large language model required for the basic version of the intelligent agent to be constructed, the identifier of at least one second data source, and the target model parameters; collecting the original second data set corresponding to the basic version of the intelligent agent from at least one second data source according to the identifier of at least one second data source; and calling the data service to perform data preprocessing on at least one original second data set to obtain at least one second data set; selecting a target open source large language model from multiple open source large language models according to the identifier of the target open source large language model; and using at least one second data set to perform model training on the target open source large language model according to the work processing flow of the basic version of the intelligent agent until the target open source large language model has the target model parameters, and using the target open source large language model with the target model parameters as the basic version of the intelligent agent.

[0043] Specifically, it supports flexible creation of various basic intelligent agents. For the basic intelligent agent to be created, first, the user can easily arrange the definition information of the work process of the basic intelligent agent on the graphical process arrangement interface, and call the process arrangement service to arrange the work process of the basic intelligent agent according to the definition information of the work process. Among them, the process arrangement service is a service used to manage and coordinate the workflow between multiple services or tasks.

[0044] For example, the work process of the intelligent customer service agent is as follows: 1. Receive user requests. 2. Problem classification, such as "account problem", "order inquiry", "technical support". 3. Data collection: Collect necessary information according to the type of problem, such as user account, order number, etc. 4. Problem solving: Provide solutions according to the type of problem, such as sending links, performing operations, etc. 5. Transfer to manual customer service: If the problem is complicated, transfer to the manual customer service team. 6. End the conversation: Send a message to end the conversation, such as "Thank you for your consultation, goodbye!"

[0045] Secondly, after creating a work process for the basic version of the intelligent agent to be created, the user performs a creation operation on the graphical intelligent agent creation interface, and configures the configuration information of the basic version of the intelligent agent to be constructed on the intelligent agent creation interface. The configuration information is used to clarify which open source large language model to use, which data sources to use or the target model parameters required for model training, etc. Optionally, the configuration information includes, for example, but is not limited to: the identifier of the target open source large language model required for the basic version of the intelligent agent to be constructed, the identifier of at least one second data source, and the target model parameters.

[0046] Common open source large language models include Qwen, GPT-3, Llama, etc. Among them, Qwen is an ultra-large-scale language model, which is trained based on a large amount of Internet text and aims to understand and generate natural language. It can be used for a variety of tasks such as text generation, question and answer, and dialogue. Qwen performs well in multiple Chinese natural language processing tasks and is suitable for applications that need to process Chinese scenes. GPT-3 (Generative Pre-trained Transformer 3) is an advanced language model known for its large scale and powerful natural language processing capabilities. GPT-3 has more than 175 billion parameters, which enables it to demonstrate excellent capabilities in a variety of natural language tasks. From text generation, translation, question and answer to code writing, GPT-3 is able to handle a very wide range of task types. Even without dedicated training data, GPT-3 is able to understand and perform tasks through simple instructions, which greatly reduces the threshold for use and achieves zero-sample learning. Llama is a series of large-scale pre-trained language models designed for a wide range of natural language processing tasks. It is a multi-scale model and supports multiple languages.

[0047] In this embodiment, the second data source can be any data source that can provide additional information or data to enhance the function of the intelligent agent. For example, it can be a knowledge graph, a database, or a data set in a specific field. Selecting a suitable second data source helps to improve the professionalism and accuracy of the intelligent agent.

[0048] Optionally, target model parameters include, but are not limited to, learning rate, batch size, training epochs, etc. These parameters directly affect the effect and efficiency of model training.

[0049] In this embodiment, when creating a basic version of the intelligent agent, the original second data set corresponding to the basic version of the intelligent agent is collected from at least one second data source, and the original second data set is also a collection of data collected from at least one second data source. The data service is called to perform data preprocessing on at least one original second data set to obtain at least one second data set, so that the at least one second data set is more suitable for model training. Among them, the data service can provide various data processing operations such as data cleaning, conversion, standardization, and integration.

[0050] In this embodiment, after obtaining at least one second data set, a target open source large language model is selected from multiple open source large language models according to the identifier of the target open source large language model; and the target open source large language model is trained using at least one second data set according to the working process of the basic version of the intelligent agent until the target open source large language model has target model parameters, and the target open source large language model with the target model parameters is used as the basic version of the intelligent agent.

[0051] It is understandable that during the model training process, the target open source large language model performs tasks according to the work process of the basic version of the intelligent agent, and simulates the work process of the intelligent agent during the training process. The purpose of this is to ensure that the target open source large language model can better adapt to actual application scenarios during training.

[0052] In some optional embodiments, after calling at least one dedicated intelligent agent to provide brand operation services to the brand operator, it is also possible to: periodically obtain relevant feedback information on the online operation effect of at least one dedicated intelligent agent from multiple channels; and evaluate the satisfaction of at least one dedicated intelligent agent in each cycle based on the relevant feedback information in the cycle; if the satisfaction of at least one dedicated intelligent agent does not meet the requirements for multiple consecutive cycles, optimize the model parameters of at least one dedicated intelligent agent.

[0053] For example, relevant feedback information can be obtained from the following channels: user feedback can be collected through evaluations or complaints submitted by users; relevant feedback information can be extracted from the log files of intelligent agents; user behavior information can be obtained by calling third-party services, etc., but not limited to these.

[0054] It can be understood that periodically obtaining feedback information can monitor the operating effect of the intelligent agent in real time, and multi-channel feedback can ensure the comprehensiveness and accuracy of the satisfaction evaluation of the dedicated intelligent agent.

[0055] In practical applications, big data mining technology can be used to evaluate the satisfaction of at least one dedicated intelligent agent in each cycle based on the relevant feedback information in each cycle, but it is not limited to this. Further optionally, in order to improve the accuracy of the evaluation, the implementation method of evaluating the satisfaction of at least one dedicated intelligent agent in each cycle based on the relevant feedback information in each cycle is: inputting the relevant feedback information in each cycle into a pre-trained satisfaction evaluation model for evaluation, and obtaining the satisfaction of at least one dedicated intelligent agent in the cycle.

[0056] Among them, the satisfaction evaluation model is a machine learning model that is trained to evaluate the satisfaction of a dedicated intelligent agent based on the relevant feedback information of the dedicated intelligent agent. During model training, relevant feedback information of multiple dedicated intelligent agents is collected as sample relevant feedback information, and the satisfaction of the sample relevant feedback information is annotated. The satisfaction levels are as follows from high to low: very satisfied, generally satisfied, and very dissatisfied. Alternatively, satisfaction is scored, for example, 90 points, 100 points, 20 points, and so on. During specific model training, the sample relevant feedback information is input into the satisfaction evaluation model to obtain the satisfaction evaluated by the satisfaction evaluation model, calculate the loss function based on the evaluated satisfaction and the annotated satisfaction, adjust the model parameters of the satisfaction evaluation model based on the loss function, and iterate the above model training process until the satisfaction evaluation model converges.

[0057] In this embodiment, the long-term performance of the intelligent agent can be more accurately judged by evaluating the satisfaction of multiple consecutive cycles. In practical applications, the multiple consecutive cycles can be flexibly set as needed, such as 5 cycles, etc. The satisfaction threshold can also be set as needed. If the satisfaction of multiple consecutive cycles is lower than the satisfaction threshold, the optimization mechanism is triggered. With the goal of optimizing the model parameters of the dedicated intelligent agent, the dedicated intelligent agent is trained to improve the model performance of the dedicated intelligent agent.

[0058] It is understandable that by periodically obtaining feedback information from multiple channels and monitoring the operating results of the intelligent agent in real time, once a decline in satisfaction is found, a quick response can be made to continuously optimize the performance of the intelligent agent and improve user satisfaction.

[0059] In some optional embodiments, the brand operator may also require that the model of the first basic version intelligent agent selected from the multiple basic version intelligent agents be optimized to obtain a more satisfactory dedicated intelligent agent. Figure 2 Make an introduction. Figure 2 A flowchart of a brand operation method provided in an embodiment of the present application. Figure 2 , the method may include the following steps:

[0060] 201. Obtain smart agent demand information of a brand operator, wherein the smart agent demand information includes at least one smart agent type required by the brand operator, working parameters corresponding to each smart agent type, an identifier of at least one first data source, and optimized model parameters.

[0061] In this embodiment, the first data source can be any data source that can provide additional information or data to enhance the intelligent agent function, for example, it can be a knowledge graph, a database, or a dataset in a specific field. The optimized model parameters can be flexibly set by the brand operator as needed, for example, including but not limited to: learning rate, batch size, training rounds, etc.

[0062] 202. Select at least one first basic version intelligent agent from a plurality of basic version intelligent agents that matches at least one intelligent agent type required by a brand operator.

[0063] 203. For the first basic version of the intelligent agent corresponding to each intelligent agent type, collect at least one original first data set corresponding to the brand operator and the intelligent agent type from at least one first data source according to the identifier of at least one first data source; and call the data service to perform data preprocessing on the at least one original first data set to obtain at least one first data set corresponding to the brand operator and the intelligent agent type; use the at least one first data set to fine-tune the first basic version of the intelligent agent until the first basic version of the intelligent agent has optimized model parameters.

[0064] In this embodiment, the original first data set is a set of data collected from at least one first data source. The method of preprocessing the original first data set to obtain the first data set can refer to the method of preprocessing the original second data set to obtain the second data set, which will not be repeated here.

[0065] It is understandable that the first basic version of the intelligent agent can be understood as a pre-trained model, which has been trained on large-scale data and has good generalization performance. Fine-tuning the first basic version of the intelligent agent using at least one first data set can make the specialized intelligent agent obtained by fine-tuning the first basic version of the intelligent agent better adapt to specific brand operation scenarios, better understand brand-specific contextual information, and provide more personalized and accurate responses.

[0066] 204. According to the working parameters corresponding to at least one intelligent agent type required by the brand operator, respectively configure at least one first basic version intelligent agent to obtain at least one dedicated intelligent agent;

[0067] 205. Call at least one dedicated intelligent agent to provide brand operation services for the brand operator.

[0068] For more information about step 201, step 202, step 204, step 205, etc., please refer to the aforementioned embodiment, which will not be repeated here.

[0069] The technical solution provided in the embodiment of the present application obtains the smart agent demand information of the brand operator, selects the smart agent that matches the demand from multiple basic version smart agents, and ensures that the selection of the smart agent is targeted and applicable; according to the smart agent demand information of the brand operator, the selected basic version smart agent is configured in detail to obtain a dedicated smart agent that can meet the specific needs of the brand operator. In particular, the dedicated smart agent has been fine-tuned and trained to better adapt to specific brand operation scenarios, can better understand brand-specific contextual information, and provide more personalized and accurate responses. Dedicated smart agents can provide highly personalized services, meet the specific needs of brand operators, improve service quality and customer satisfaction, improve the efficiency and quality of brand operations, significantly improve the overall level of brand operations, and enhance the brand's market competitiveness.

[0070] For a better understanding, the following Figure 3a The full-link intelligent body base for brand operation shown in the figure is introduced. The full-link intelligent body base for brand operation can be understood as the executive entity of brand operation, which can be composed of software and / or hardware. The full-link intelligent body base for brand operation provides multiple intelligent bodies for brand operators to choose from. Multiple intelligent bodies include, but are not limited to: intelligent customer service agent, legal assistant agent, network assistant agent and intelligent shopping guide agent. The full-link intelligent body base for brand operation provides data services and supports open source large models. The full-link intelligent body base for brand operation also has certain basic AI (Artificial Intelligence) capabilities, supports the collaborative work of multiple intelligent agents, etc.

[0071] Figure 3b This is a schematic diagram of the structure of a brand operation device provided in an embodiment of the present application. The device may be composed of software and / or hardware and may generally be integrated into an electronic device, see Figure 3b , the device may include:

[0072] An acquisition module 31 is used to acquire the smart agent demand information of the brand operator, wherein the smart agent demand information includes at least one smart agent type required by the brand operator and the working parameters corresponding to each smart agent type;

[0073] A selection module 32, configured to select at least one first basic version intelligent agent matching at least one intelligent agent type required by a brand operator from among a plurality of basic version intelligent agents;

[0074] A configuration module 33, configured to configure at least one first basic version intelligent agent according to working parameters corresponding to at least one intelligent agent type required by the brand operator, to obtain at least one dedicated intelligent agent;

[0075] The calling module 34 is used to call at least one dedicated intelligent agent to provide brand operation services for the brand operator.

[0076] Further optionally, the intelligent agent requirement information further includes: an identifier of at least one first data source corresponding to each intelligent agent type and optimized model parameters;

[0077] Correspondingly, the above-mentioned device also includes: a fine-tuning training module, which is used to collect at least one original first data set corresponding to the brand operator and the intelligent agent type from at least one first data source according to the identifier of at least one first data source for the first basic version of the intelligent agent corresponding to each intelligent agent type; and call the data service to perform data preprocessing on the at least one original first data set to obtain at least one first data set corresponding to the brand operator and the intelligent agent type; use the at least one first data set to fine-tune the first basic version of the intelligent agent until the first basic version of the intelligent agent has optimized model parameters.

[0078] Further optionally, the above-mentioned device also includes: a creation module, which is used to: in response to a process orchestration operation for any basic version of the intelligent agent to be built on the process orchestration interface, call the process orchestration service to orchestrate the work processing flow of the basic version of the intelligent agent; in response to the creation operation initiated on the intelligent agent creation interface, obtain the configuration information of the basic version of the intelligent agent to be built, the configuration information including the identifier of the target open source large language model required for the basic version of the intelligent agent to be built, the identifier of at least one second data source and the target model parameters; collect the original second data set corresponding to the basic version of the intelligent agent from at least one second data source according to the identifier of at least one second data source; and call the data service to perform data preprocessing on at least one original second data set to obtain at least one second data set; select the target open source large language model from multiple open source large language models according to the identifier of the target open source large language model; and use at least one second data set to perform model training on the target open source large language model according to the work processing flow of the basic version of the intelligent agent until the target open source large language model has the target model parameters, and use the target open source large language model with the target model parameters as the basic version of the intelligent agent.

[0079] Further optionally, the dedicated intelligent agent includes at least one of the following: an intelligent customer service agent Agent, a legal assistant Agent, a networking assistant Agent, and an intelligent shopping guide Agent.

[0080] Further optionally, the above-mentioned device also includes: an optimization module, which is used to periodically obtain relevant feedback information on the online operation effect of at least one dedicated intelligent agent from multiple channels; and evaluate the satisfaction of at least one dedicated intelligent agent in each cycle based on the relevant feedback information in the cycle; if the satisfaction of at least one dedicated intelligent agent does not meet the requirements for multiple consecutive cycles, then optimize the model parameters of at least one dedicated intelligent agent.

[0081] Further optionally, when the optimization module evaluates the satisfaction level of at least one dedicated intelligent agent in each cycle based on the relevant feedback information in that cycle, it is specifically used to: input the relevant feedback information in each cycle into a pre-trained satisfaction evaluation model for evaluation to obtain the satisfaction level of at least one dedicated intelligent agent in that cycle.

[0082] Figure 3b The device shown can perform Figure 1 or Figure 2 The method shown in the embodiment shown, its implementation principle and technical effect are not described in detail. Figure 3b The specific manner in which each module and unit of the device shown performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0083] It should be noted that the execution subject of each step of the method provided in the above embodiment can be the same device, or the method can be executed by different devices. For example, the execution subject of steps 101 to 104 can be device A; for another example, the execution subject of steps 101 and 102 can be device A, and the execution subject of steps 103 and 104 can be device B; and so on.

[0084] In addition, in some of the processes described in the above embodiments and the accompanying drawings, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel, and the sequence numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the sequence numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0085] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0086] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 4 As shown, the electronic device includes: a memory 41 and a processor 42;

[0087] The memory 41 is used to store computer programs and can be configured to store various other data to support operations on the computing platform. Examples of such data include instructions for any application or method operating on the computing platform, contact data, phone book data, messages, pictures, videos, etc.

[0088] The memory 41 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read only memory (EEPROM), erasable programmable read only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0089] The processor 42 is coupled to the memory 41 and is used to execute the computer program in the memory 41 to execute the steps in the brand operation method.

[0090] Further optional, such as Figure 4 As shown, the electronic device also includes: a communication component 43, a display 44, a power component 45, an audio component 46 and other components. Figure 4 Only some components are shown schematically, which does not mean that the electronic device only includes Figure 4 In addition, Figure 4The components in the dashed box are optional components, not mandatory components, and the specific components depend on the product form of the electronic device. The electronic device of this embodiment can be implemented as a terminal device such as a desktop computer, a laptop computer, a smart phone, or an IOT (Internet of Things) device, or a server device such as a conventional server, a cloud server, or a server array. If the electronic device of this embodiment is implemented as a terminal device such as a desktop computer, a laptop computer, a smart phone, etc., it can include Figure 4 If the electronic device of this embodiment is implemented as a server device such as a conventional server, a cloud server or a server array, it may not include Figure 4 Components within the dashed box.

[0091] For the detailed implementation process of the processor executing each action, please refer to the relevant description in the aforementioned method embodiment or device embodiment, which will not be repeated here.

[0092] Accordingly, an embodiment of the present application further provides a computer-readable storage medium storing a computer program, which, when executed, can implement each step that can be executed by an electronic device in the above method embodiment.

[0093] Accordingly, an embodiment of the present application also provides a computer program product, including a computer program / instruction. When the computer program / instruction is executed by a processor, the processor is enabled to implement each step in the above method embodiment that can be executed by an electronic device.

[0094] The above-mentioned communication component is configured to facilitate wired or wireless communication between the device where the communication component is located and other devices. The device where the communication component is located can access a wireless network based on a communication standard, such as WiFi (Wireless Fidelity), 2G (2 Generation, 2 Generation), 3G (3 Generation, 3 Generation), 4G (4 Generation, 4 Generation) / LTE (long Term Evolution), 5G (5 Generation, 5 Generation) and other mobile communication networks, or a combination thereof. In an exemplary embodiment, the communication component receives a broadcast signal or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component also includes a near field communication (Near Field Communication, NFC) module to facilitate short-range communication. For example, the NFC module can be based on Radio Frequency Identification (RFID) technology, Infrared Data Association (IrDA) technology, Ultra Wide Band (UWB) technology, Bluetooth (BT) technology and other technologies.

[0095] The above-mentioned display includes a screen, and the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from a user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure associated with the touch or slide operation.

[0096] The power supply assembly provides power to various components of the device where the power supply assembly is located. The power supply assembly may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to the device where the power supply assembly is located.

[0097] The above-mentioned audio component can be configured to output and / or input audio signals. For example, the audio component includes a microphone (MIC), and when the device where the audio component is located is in an operating mode, such as a call mode, a recording mode, and a speech recognition mode, the microphone is configured to receive an external audio signal. The received audio signal can be further stored in a memory or sent via a communication component. In some embodiments, the audio component also includes a speaker for outputting an audio signal.

[0098] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-readable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0099] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0100] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0101] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0102] In a typical configuration, a computing device includes one or more processors (Central Processing Unit, CPU), input / output interface, network interface and memory.

[0103] Memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0104] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, Phase Change RAM (PRAM), Static Random-Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically-Erasable Programmable Read-Only Memory (EEPROM), flash memory or other memory technology, CD-ROM, Digital Versatile Disc (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. According to the definition in this article, computer-readable media does not include temporary computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0105] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0106] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A brand operation method, characterized in that: include: Acquire intelligent agent demand information of a brand operator, wherein the intelligent agent demand information includes at least one intelligent agent type required by the brand operator and working parameters corresponding to each intelligent agent type; Selecting at least one first basic version intelligent agent that matches at least one intelligent agent type required by the brand operator from among multiple basic version intelligent agents; According to the working parameters corresponding to at least one intelligent agent type required by the brand operator, respectively configure at least one first basic version intelligent agent to obtain at least one dedicated intelligent agent; At least one dedicated intelligent agent is called to provide brand operation services for the brand operator.

2. The method according to claim 1, characterized in that The intelligent agent requirement information also includes: an identifier of at least one first data source corresponding to each intelligent agent type and optimized model parameters; Accordingly, according to the working parameters corresponding to at least one intelligent agent type required by the brand operator, at least one first basic version intelligent agent is configured respectively to obtain at least one dedicated intelligent agent, and the method further includes: For the first basic version of the intelligent agent corresponding to each intelligent agent type, at least one original first data set corresponding to the brand operator and the intelligent agent type is collected from at least one first data source according to the identifier of at least one first data source; and a data service is called to perform data preprocessing on the at least one original first data set to obtain at least one first data set corresponding to the brand operator and the intelligent agent type; The first basic version intelligent agent is fine-tuned and trained using the at least one first data set until the first basic version intelligent agent has optimized model parameters.

3. The method according to claim 1, characterized in that Before obtaining the brand operator's intelligent agent demand information, it also includes: In response to a process orchestration operation on a process orchestration interface for any basic version intelligent agent to be constructed, calling a process orchestration service to orchestrate a work processing flow of the basic version intelligent agent; In response to a creation operation initiated on the intelligent agent creation interface, configuration information of a basic version of the intelligent agent to be constructed is obtained, wherein the configuration information includes an identifier of a target open source large language model required by the basic version of the intelligent agent to be constructed, an identifier of at least one second data source, and target model parameters; According to the identifier of at least one second data source, an original second data set corresponding to the basic version intelligent agent is collected from at least one second data source; and a data service is called to perform data preprocessing on at least one original second data set to obtain at least one second data set; The target open source large language model is selected from multiple open source large language models according to the identifier of the target open source large language model; and the target open source large language model is trained using the at least one second data set according to the working process of the basic version intelligent agent until the target open source large language model has the target model parameters, and the target open source large language model with the target model parameters is used as the basic version intelligent agent.

4. The method according to claim 1, characterized in that: The dedicated intelligent agent includes at least one of the following: an intelligent customer service agent Agent, a legal assistant Agent, a networking assistant Agent and an intelligent shopping guide Agent.

5. The method according to claim 1, characterized in that After calling at least one dedicated intelligent agent to provide brand operation services for the brand operator, the method further includes: Periodically obtaining relevant feedback information on the online operation effect of at least one dedicated intelligent agent from multiple channels; and evaluating the satisfaction of the at least one dedicated intelligent agent in each cycle based on the relevant feedback information in the cycle; If the satisfaction level of at least one dedicated intelligent agent does not meet the requirements for multiple consecutive cycles, the model parameters of at least one dedicated intelligent agent are optimized.

6. The method according to claim 5, characterized in that Evaluating the satisfaction of the at least one dedicated intelligent agent in each cycle according to relevant feedback information in the cycle, including: The relevant feedback information in each cycle is input into a pre-trained satisfaction evaluation model for evaluation to obtain the satisfaction of the at least one dedicated intelligent agent in the cycle.

7. A brand operation device, characterized in that: include: An acquisition module, used to acquire the smart agent demand information of the brand operator, wherein the smart agent demand information includes at least one smart agent type required by the brand operator and working parameters corresponding to each smart agent type; A selection module, configured to select at least one first basic version intelligent agent matching at least one intelligent agent type required by the brand operator from among a plurality of basic version intelligent agents; A configuration module, configured to configure at least one first basic version intelligent agent according to working parameters corresponding to at least one intelligent agent type required by the brand operator, to obtain at least one dedicated intelligent agent; The calling module is used to call at least one dedicated intelligent agent to provide brand operation services for the brand operator.

8. An electronic device, characterized in that: include: Memory and processor; The memory is used to store computer programs; The processor is coupled to the memory and configured to execute the computer program to perform the steps of the method according to any one of claims 1 to 7.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the processor is enabled to implement the steps in the method according to any one of claims 1 to 7.

10. A computer program product, characterized in that The method comprises a computer program / instruction, which, when executed by a processor, enables the processor to implement the steps of the method according to any one of claims 1 to 7.

Images