Product promotion strategy generation method and device, equipment and medium

Abstract

The embodiment of the invention provides a product promotion strategy generation method and device, equipment and a medium. The scheme comprises the steps of obtaining a request provided by a user to generate initial request information of a promotion strategy for a target product; the first agent processes the initial request information to obtain multiple pieces of prompt word information for multiple dimensions, and the multiple dimensions at least comprise one of the user dimension and the product dimension; and each second agent generates a corresponding product subdivision promotion strategy according to the prompt word information which can be solved by the second agent, and the first agent integrates the product subdivision promotion strategies to obtain a product promotion strategy corresponding to the initial request information.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method, device, equipment, and medium for generating a product promotion strategy. Background Art

[0002] With the continuous advancement of science and technology, a wide variety of products are emerging. To help users more accurately understand product information and enhance their loyalty, it is crucial to develop effective and scientific product promotion strategies. Precise promotional methods not only enhance product competitiveness but also strengthen user engagement, encouraging them to better understand the product's unique value and advantages, thereby increasing product usage and brand loyalty.

[0003] Currently, product promotion strategies are mainly formulated manually by operations personnel. Manual formulation often takes a relatively long time, which is not conducive to the rapid layout of products in the rapidly changing market.

[0004] Therefore, an efficient method for generating product promotion strategies is needed. Summary of the Invention

[0005] The embodiments of this specification provide a method, apparatus, device, and medium for generating a product promotion strategy to improve the efficiency of generating a product promotion strategy.

[0006] To solve the above technical problems, the embodiments of this specification are implemented as follows.

[0007] In a first aspect, embodiments of this specification provide a method for generating a product promotion strategy. The method is applied to a product promotion strategy generation system. The product promotion strategy generation system includes a first agent for generating multiple prompt word information in multiple dimensions based on a user's initial request information, and a second agent for generating a product promotion strategy for each dimension of prompt word information. The method includes:

[0008] Obtaining the request provided by the user to generate initial request information for a promotion strategy for a target product;

[0009] The first agent processes the initial request information to obtain a plurality of prompt word information for a plurality of dimensions; different prompt word information is used to describe the request for generating a promotion strategy for the target product from different dimensions; the plurality of dimensions include at least one of a user dimension and a product dimension;

[0010] Utilizing each of the second intelligent agents, generating a product segmentation promotion strategy corresponding to each of the prompt word information;

[0011] Based on each of the product segmentation promotion strategies generated by each of the second intelligent agents, the product promotion strategy corresponding to the initial request information is obtained.

[0012] In a second aspect, embodiments of this specification provide a device for generating a product promotion strategy. The device is applied to a product promotion strategy generation system. The product promotion strategy generation system includes a first agent for generating multiple prompt word information in multiple dimensions based on a user's initial request information, and a second agent for generating a product promotion strategy for each dimension of prompt word information. The second agent includes:

[0013] An initial request information acquisition module is used to obtain initial request information provided by a user to generate a promotion strategy for a target product;

[0014] an initial request information processing module, configured to process the initial request information using the first agent to obtain a plurality of prompt word information for a plurality of dimensions; different prompt word information is used to describe a request for generating a promotion strategy for the target product from different dimensions; the plurality of dimensions including at least one of a user dimension and a product dimension;

[0015] A product segmentation promotion strategy generation module, configured to generate a product segmentation promotion strategy corresponding to each of the prompt word information using each of the second agents;

[0016] The product promotion strategy obtaining module is used to obtain the product promotion strategy corresponding to the initial request information based on each of the product segmentation promotion strategies generated by each of the second intelligent agents.

[0017] In a third aspect, an embodiment of this specification provides a device for controlling a smart container, including:

[0018] at least one processor; and,

[0019] a memory communicatively connected to the at least one processor; wherein,

[0020] The memory stores instructions that can be executed by the at least one processor. The instructions are executed by the at least one processor to enable the at least one processor to execute the above-mentioned product promotion strategy generation method.

[0021] In a fourth aspect, an embodiment of this specification provides a computer-readable medium having computer-readable instructions stored thereon, and the computer-readable instructions can be executed by a processor to implement the above-mentioned product promotion strategy generation method.

[0022] At least one embodiment of this specification can achieve the following beneficial effects:

[0023] The first intelligent agent generates initial request information for the promotion strategy of the target product from the request provided by the user, and decomposes it into multiple prompt word information of multiple dimensions. Then, each second intelligent agent processes the multiple prompt word information of multiple dimensions to obtain the product segmentation promotion strategy corresponding to each prompt word information. Finally, based on each product segmentation promotion strategy, a product promotion strategy corresponding to the initial request information is generated. In this way, the product promotion strategy for the initial request information can be automatically generated based on the first intelligent agent and the second intelligent agent to improve the efficiency of generating the product promotion strategy.

[0024] The prompt word information processed by each second intelligent agent is generated by the first intelligent agent based on the initial request information provided by the user, so that each second intelligent agent can process the request information provided by the user under the guidance of the first intelligent agent, improve the consistency between the product segmentation promotion strategy generated by each second intelligent agent and the generation result expected by the first intelligent agent, and avoid the divergence of the request information in the process of processing the request information provided by the user, so as to improve the accuracy of the product promotion strategy generated for the initial request information provided by the user. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the embodiments of this specification or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments recorded in this application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0026] Figure 1 This is a schematic diagram of an application scenario of a product promotion strategy generation method provided in an embodiment of this specification;

[0027] Figure 2 This is a flowchart of a method for generating a product promotion strategy provided by an embodiment of this specification;

[0028] Figure 3 This is a schematic diagram of a hierarchical multi-agent system provided in this manual;

[0029] Figure 4 This is a swim lane diagram of a method for generating a product promotion strategy provided in an embodiment of this specification;

[0030] Figure 5 This specification provides the corresponding embodiment Figure 2 A structural diagram of a device for generating a product promotion strategy;

[0031] Figure 6 This is a structural diagram of a product promotion strategy generation device provided in an embodiment of this specification. DETAILED DESCRIPTION

[0032] To make the purpose, technical solutions, and advantages of one or more embodiments of this specification more clear, the technical solutions of one or more embodiments of this specification will be clearly and completely described below in conjunction with the specific embodiments of this specification and the corresponding drawings. Obviously, the embodiments described are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of one or more embodiments of this specification.

[0033] To facilitate understanding of this specification, at least some of the terms in this specification are explained as follows:

[0034] An agent, or AI agent model, is a highly intelligent model, distinct from traditional machine learning models. It can perceive its environment, make decisions, and execute tasks. The core of the AI ​​agent model lies in its ability to think independently and utilize tools, enabling it to gradually achieve a given goal. During the use of the agent model, it can gradually complete a specified task based on predefined roles and path planning.

[0035] A product promotion strategy is a series of plans and actions a company takes to increase product awareness, increase market share, attract potential customers, and facilitate sales. These strategies can help companies effectively communicate their product's value proposition and connect with target customers.

[0036] Prompt word information usually refers to the specific guidance or instructive content given when interacting with the large language model. It is used to help the large language model understand the user's needs, questions or tasks. Prompt word information can help the large language model better analyze the user's intent to ensure that it can provide appropriate and accurate responses.

[0037] A multi-agent system is a computing system composed of multiple interacting agents that can autonomously perform tasks, solve problems, and achieve the overall goals of the system or solve complex problems through cooperation or competition.

[0038] In existing multi-agent systems, when processing user-provided request information, each agent operates in a leaderless state, solving problems through inter-agent dialogue. This approach, due to the lack of leadership among agents, allows each agent to propose solutions independently. This can lead to disagreements between agents, making it easy for multiple rounds of dialogue to produce a final result. Even if the multi-agent system does produce a result, it can often diverge from the user's needs.

[0039] In order to solve the defects in the related art, in the embodiments of this specification, the first intelligent agent processes the request information provided by the user to obtain multiple prompt word information of multiple dimensions, and each second intelligent agent selects the prompt word information that it can solve from the multiple prompt word information of multiple dimensions for processing, so that each second intelligent agent can process the request information under the leadership of the first intelligent agent, so as to achieve the unity of opinions of each second intelligent agent and the first intelligent agent on solving the request information, so as to avoid divergence problems among multiple intelligent agents.

[0040] In addition, in the embodiments of the present specification, during the process of processing the request information, the available second intelligent agent can process the prompt words of one dimension among multiple prompt word information of multiple dimensions, and the prompt word information of one dimension is processed by one of the second intelligent agents. Different second intelligent agents process different prompt word information. After the available second intelligent agents process the prompt word information that they can solve respectively, they can obtain the product promotion strategy for the request information based on the generated product segmentation promotion strategies, so as to improve the generation efficiency of the product promotion strategy.

[0041] The technical solutions provided by the embodiments of this specification are described in detail below with reference to the accompanying drawings. Figure 1 This is a schematic diagram of an application scenario of a product promotion strategy generation method provided in an embodiment of this specification.

[0042] like Figure 1 As shown, the application scenario diagram includes a user terminal 101 and a server 102.

[0043] In the embodiments of this specification, the user terminal 101 may include but is not limited to at least one of a smart phone, a tablet computer, a laptop computer, an intelligent interactive device, a wearable device, an in-vehicle smart terminal, etc., wherein the wearable device may include but is not limited to: a smart bracelet, a smart watch, smart glasses, etc.

[0044] The server 102 may include but is not limited to at least one of any device, equipment, platform, equipment cluster, or cloud computing service center with computing and processing capabilities.

[0045] A communication connection is established between the user terminal 101 and the server 102, wherein the communication connection method may include but is not limited to a local area network connection, a wide area network connection, an Internet connection, a short-range communication connection, or other types of data network connections. Short-range communication connections include but are not limited to near field communication (NFC), local area network, Bluetooth, infrared, and other connection methods.

[0046] A user using the user terminal 101 can send a request message to the server 102 through the user terminal 101 to request the generation of a promotion strategy for a target product. In response to the request message sent by the user terminal 101, the server 102 can call the first agent to process the request message to obtain multiple prompt word information in multiple dimensions, call the second agent to generate a product segmentation promotion strategy for the prompt word information in each dimension, and call the first agent to integrate the various product segmentation promotion strategies generated by the second agent into a product promotion strategy.

[0047] In actual applications, if the memory resources of the user terminal 101 are sufficient to support the operation of the agent, the process corresponding to each agent can also be executed on the user terminal 101.

[0048] Next, a method for generating a product promotion strategy provided in an embodiment of the specification will be described with reference to the accompanying drawings.

[0049] Figure 2 This is a flowchart of a method for generating a product promotion strategy provided in an embodiment of this specification. From a program perspective, the execution body of the process can be a program installed on a server used to generate the product promotion strategy. From a hardware perspective, the execution body of the process can be a device used to generate the product promotion strategy.

[0050] like Figure 2 As shown, the process may include the following steps:

[0051] Step 202: Obtain initial request information provided by the user to generate a promotion strategy for the target product.

[0052] In the embodiments of this specification, target products may include tangible physical products that can be touched, such as cars, furniture, clothes, books, food, etc. Target products may also include virtual, intangible digital products, typically used through electronic devices, such as mobile applications, e-books, online courses, games, software tools, etc. In e-commerce application scenarios, the target product may refer to an installment payment service used to pay for commodity transactions in the e-commerce scenario.

[0053] Promotion strategy refers to a strategic plan developed for a target product to increase product awareness, increase market share, attract potential customers and facilitate sales.

[0054] The initial request information may be request information provided by a user to a target device during interaction with the target device used to generate a product promotion strategy. Alternatively, the initial request information may be request information pre-saved by the user in a target database. The device used to generate a product promotion strategy may access the target database and retrieve the pre-saved request information from the target database.

[0055] The user may refer to the operations personnel who are responsible for providing product promotion strategies for the target product, or may refer to the seller of goods involved in the target product. Involving the target product may refer to the act of selling the target product, or may refer to the act of using the target product in the process of selling other goods. For example, if the other goods sold are clothing, the target product may be an installment payment service used to pay for the clothing.

[0056] For example, if the target product is an installment payment service, the initial request information for the target product may be "What promotion strategies are there to encourage users to use installment payment?"

[0057] Step 204: Utilize the first agent to process the initial request information to obtain multiple prompt word information for multiple dimensions; different prompt word information is used to describe the request for generating a promotion strategy for the target product from different dimensions; the multiple dimensions include at least one dimension from the user dimension and the product dimension.

[0058] In the embodiments of this specification, the first agent may be the main agent that processes the initial request information. Based on the initial request information, the first agent may generate multiple prompt word information in multiple dimensions. Different prompt word information may describe the request from different dimensions to generate a promotion strategy for the target product. In practical applications, the multiple dimensions may include any one or more of the following: user dimension, product dimension, quota dimension, and supply dimension.

[0059] In the embodiments of this specification, multiple dimensions may represent analysis or description of the user's request information from different angles or aspects, and each dimension may represent a specific perspective, through which the user's needs may be fully understood or answered, so as to ensure that various factors in the request information are fully considered, so as to provide the user with a complete and effective answer.

[0060] Multiple prompt word information in multiple dimensions may refer to one prompt word information for one dimension, or may refer to multiple prompt word information for one dimension, which is not limited to this.

[0061] Continuing with the above example, assuming that the initial request information provided by the user is "Are there any promotional strategies for encouraging users to pay in installments?", the initial request information is processed, and the multiple prompt word information of multiple dimensions obtained may include the first prompt word information "Are there any promotional strategies for user management for encouraging users to pay in installments?", the second prompt word information "Are there any promotional strategies for product management for encouraging users to pay in installments?", the third prompt word information "Are there any promotional strategies for users' credit limits for encouraging users to pay in installments?", and the fourth prompt word information "Are there any promotional strategies for suppliers for encouraging users to pay in installments?", etc.

[0062] Step 206: Utilize each of the second intelligent agents to generate a product segmentation promotion strategy corresponding to each of the prompt word information.

[0063] In the embodiments of this specification, the second agent may be a sub-agent that processes the initial request information. A hierarchical multi-agent system comprising multiple agents may be formed based on the first agent and the second agent, wherein the hierarchical multi-agent system may include a top layer, a middle layer, and a bottom layer, wherein the top layer may be provided with a first agent, the middle layer may be provided with a second agent, and the bottom layer may provide tools and resources to support the operation and collaboration of the first agent and the second agent. The tools may be a set of technical components called by the first agent or the second agent to achieve a specific function, such as tools for creating user tables, tools for outputting strategies, tools for exploring market opportunities, etc. Resources may be supporting factors used to assist the first agent or the second agent in achieving a certain function, such as computing resources, storage resources, data resources, software resources, hardware resources, etc.

[0064] In a multi-agent system based on a hierarchical model consisting of a first agent and a second agent, the first agent can provide multiple prompt word information generated in multiple dimensions to the second agent. Specifically, the first agent can send the multiple prompt word information generated in multiple dimensions to the bottom layer, and the second agent can then filter the prompt word information from the bottom layer to solve the problem.

[0065] In the embodiments of this specification, a single second agent can handle prompt word information of a single dimension, while different second agents can handle prompt word information of different dimensions. Prompt word information of a single dimension can be handled by a single second agent. In practical applications, prompt word information of a single dimension can also be handled by multiple second agents, without limitation.

[0066] Continuing with the above example, if the target product is an installment payment service, the second agent may include a user operation agent for providing strategies from the user dimension, a product operation agent for providing strategies from the product dimension, a credit operation agent for providing strategies from the credit dimension, and a supply operation agent for providing strategies from the supply dimension. Among them, the user operation agent can solve the above-mentioned first prompt word information to obtain a product segmentation promotion strategy for the first prompt word information; the product operation agent can solve the above-mentioned second prompt word information to obtain a product segmentation promotion strategy for the second prompt word information; the credit operation agent can solve the above-mentioned third prompt word information to obtain a product segmentation promotion strategy for the third prompt word information; the supply operation agent can solve the above-mentioned fourth prompt word information to obtain a product segmentation promotion strategy for the fourth prompt word information. Among them, the product segmentation promotion strategy can be a promotion strategy generated for the target product in one of the multiple dimensions.

[0067] In order to facilitate those skilled in the art to understand the present solution, a specific example is provided in the embodiments of this specification for a multi-agent system in a hierarchical model.

[0068] Figure 3 This is a schematic diagram of a hierarchical multi-agent system provided in this manual. Figure 3As shown, a hierarchical multi-agent system may include a top layer 301, a middle layer 302, and a bottom layer 303. The top layer 301 may be equipped with a first agent, which may include a dedicated operation agent. The dedicated operation agent is used to plan the logical approach for processing received request information from a holistic perspective and assign local tasks for resolving the request information to the second agent in the middle layer 302. The middle layer 302 may be equipped with a second agent, which may include a user operation agent for providing operation strategies from the user dimension, a product operation agent for providing operation strategies from the product dimension, a quota operation agent for providing operation strategies from the quota dimension, and a supply operation agent for providing operation strategies from the supply dimension. The bottom layer 303 may provide tools and resources to support the operation and collaboration of the first and second agents. The tools may include strategy generation tools, market opportunity mining tools, etc., and the resources may include computing resources, storage resources, data resources, software resources, hardware resources, etc.

[0069] In order to reduce the workload of the first intelligent agent, the middle layer 302 can also be provided with a third intelligent agent, which can include a data analyst intelligent agent. The data analyst intelligent agent is used to perform data analysis on the data required to generate product promotion strategies based on the requested information, such as analyzing the target user group corresponding to the target product, and analyzing the characteristics of the target user group. Figure 3 The multi-agent system in the hierarchical model is only an example. The middle layer 302 may also include agents with other functions, or the middle layer 302 may not include any one or more of the data analyst agent, user operation agent, product operation agent, quota operation agent and supply operation agent. There is no limitation on this.

[0070] Step 208: Based on the product segmentation promotion strategies generated by each of the second intelligent agents, obtain the product promotion strategy corresponding to the initial request information.

[0071] In the embodiment of this specification, each product segmentation promotion strategy can be a promotion strategy generated from one perspective based on the initial request information provided by the user. By integrating the various product segmentation promotion strategies, a final promotion strategy generated from various perspectives based on the initial request information provided by the user can be obtained.

[0072] It should be understood that the order of some steps in the methods described in one or more embodiments of this specification can be interchanged according to actual needs, or some steps can be omitted or deleted.

[0073] Figure 2In the method, the first agent processes the initial request information provided by the user to obtain multiple prompt word information of multiple dimensions. Each second agent can process the prompt word information of one dimension of the multiple prompt word information of multiple dimensions, so that each second agent can process the request information under the leadership of the first agent, so as to achieve the unity of opinions of each second agent and the first agent on solving the request information, thereby avoiding divergence problems among multiple agents.

[0074] On the other hand, a prompt word message can be solved by only one second agent. Thus, for a prompt word message, the generated result of a second agent can be used as the product segmentation promotion strategy for the prompt word message, thereby improving the efficiency of generating product segmentation promotion strategies and avoiding the time-consuming risk caused by the simultaneous solution of the same prompt word message by multiple second agents, as each second agent will output a generated result for the prompt word message, thereby screening the product segmentation promotion strategy for the prompt word message from multiple generated results. In actual applications, a prompt word message can also be solved by multiple second agents, so that multiple product segmentation promotion strategies can be generated for the prompt word message. Thus, the promotion strategy with the greatest relevance to the prompt word message can be determined from the multiple product segmentation promotion strategies, thereby improving the diversity and accuracy of the product segmentation promotion strategies generated for the prompt word message.

[0075] based on Figure 2 The present specification also provides some specific implementation plans of the method, which are described below.

[0076] In order to further avoid the divergence problem that occurs when each intelligent agent processes the initial request information provided by the user, before processing the initial request information provided by the user, the workflow for processing the initial request information can be determined first.

[0077] Optionally, before using the first agent to process the initial request information and obtain multiple prompt word information for multiple dimensions, it may also include: in response to the initial request information, using the first agent to determine a first workflow for processing the initial request information, the first workflow is used to reflect the processing logic for processing the initial request information; based on the first workflow, using the first agent to process the initial request information.

[0078] In the embodiments of this specification, the first workflow may include a series of operations or tasks performed in a specific order and according to specific rules for processing the initial request information. The first workflow may define summary information of tasks that each agent must complete during the processing of the initial request information. The first workflow may also reflect the processing logic for processing the initial request information. The processing logic may refer to the manner in which the request information is processed according to specific rules, steps, or methods.

[0079] For example, if the initial request information for the target product is "What promotion strategies are there to encourage users to pay in installments?", the generated first workflow may include: the first agent performs intent analysis on the initial request information, the first agent calls the data analyst agent to perform data analysis, the first agent supplements the initial request information based on the data analyst's data analysis results, the first agent performs splitting processing on the request information after data supplementation, the first agent calls the second agent to process the prompt word information obtained after the splitting processing, the first agent integrates the product segmentation promotion strategies generated by each second agent, and the first agent outputs the product promotion strategy obtained after integration. For different initial request information, the content of the first workflow can be determined according to the actual scenario. The content contained in the first workflow in this specification is only an example and does not constitute a limitation on the first workflow.

[0080] In an embodiment of the present specification, before processing the initial request information, a workflow for processing the initial request information can be generated so that each intelligent agent can process the initial request information based on the workflow, thereby improving the uniformity of opinions of each intelligent agent on processing the initial request information, thereby further avoiding divergence problems between the various intelligent agents.

[0081] In the embodiments of this specification, a specific embodiment is also proposed for a method of generating multiple prompt word information in multiple dimensions.

[0082] Optionally, the use of the first intelligent agent to process the initial request information to obtain multiple prompt word information for multiple dimensions may specifically include: performing content supplementation processing on the initial request information to obtain the supplemented request information; based on the functions of each of the second intelligent agents, performing disassembly processing on the supplemented request information to obtain multiple prompt word information for the multiple dimensions; the prompt word information of one dimension at least meets the function of one of the second intelligent agents.

[0083] In the embodiments of this specification, content supplementation processing may refer to request information improvement processing that adds features to the initial request information, which is equivalent to adding known conditions to the initial request information so that when the intelligent agent processes the request information based on the content supplemented processing, it can generate a more accurate product promotion strategy.

[0084] In the embodiment of this specification, the first intelligent agent can perform content supplement processing on the initial request information, thereby eliminating the need for each second intelligent agent to perform content supplement processing separately when obtaining the prompt word information that it has solved, thereby reducing the work tasks of the second intelligent agent and improving the efficiency of the second intelligent agent in generating product segmentation promotion strategies.

[0085] In the embodiments of this specification, the first agent may also perform decomposition processing on the request information after the supplementary processing. Before decomposition processing on the request information after the supplementary processing, the first agent may traverse each second agent to obtain the target functions corresponding to the available second agents. Based on the target functions of the available second agents, the first agent may decompose the prompt word information corresponding to the target function from the request information after the supplementary processing, so that the decomposed prompt word information can be processed by the second agent having the target function.

[0086] Optionally, based on the target function of the second intelligent entity, the prompt word information corresponding to the target function is disassembled from the request information after supplementary processing, which may include: obtaining dimension label information used to reflect the target function of the second intelligent entity; adding the dimension label information to the request information after supplementary processing to obtain the prompt word information corresponding to the target function.

[0087] Optionally, the dimension label information may be keyword information extracted from text information used to describe the target function of the second agent.

[0088] For example, if the target product is an installment payment service, the request information after supplemental processing may be, "For the user group that uses installment payments, including adolescent users who are students and whose consumption level is between 0 and k yuan, what promotion strategies are there to encourage users to use installment payments?" Assuming that the second agent is a user operation agent, and the text information used to describe the target function of the user operation agent is, "to provide product promotion strategies from the user dimension," the keyword information extracted from the text information may be "user dimension," and adding this keyword information to the request information after supplemental processing may be, "For the user group that uses installment payments, including adolescent users who are students and whose consumption level is between 0 and k yuan, what promotion strategies are there to encourage users to use installment payments from the user dimension?"

[0089] In actual applications, for the multiple prompt word information of multiple dimensions decomposed from the initial request information, the supplementary content contained in any two different prompt word information relative to the initial request information may be completely the same or partially the same, and there is no limitation on this.

[0090] In the embodiment of this specification, based on the function of the second intelligent agent, prompt word information is disassembled from the request information after supplementary processing, so that the disassembled prompt word information can all be resolved by the second intelligent agent, thereby avoiding the risk of the disassembled prompt word information not being able to be processed, and improving the generation efficiency of product promotion strategies.

[0091] When performing content supplement processing on the initial request information, the supplement may be performed based on the characteristic data of the target user group corresponding to the target product, so that the generated product promotion strategy can effectively influence the target user group.

[0092] Optionally, the content supplement processing is performed on the initial request information to obtain the request information after supplement processing, which may specifically include: obtaining the characteristic data of the target user corresponding to the target product; based on the characteristic data, content supplement processing is performed on the initial request information to obtain the request information after supplement processing; the request information after supplement processing includes the characteristic data.

[0093] In the embodiments of this specification, the target user group may include at least one of the buyer and seller of the target product in the transaction process. In actual applications, the target user group can be determined based on the actual scenario and is not limited to this.

[0094] In the embodiments of this specification, feature data may include the user's life stage information, consumption level information, occupation information, seller's supply type information, scenario information involving the target product, etc., where the life stage information may include adolescence, middle age, and old age, etc.

[0095] In actual applications, the initial request information provided by the user is relatively simple, and the initial request information reflects less conditional information. When the intelligent agent generates a product promotion strategy based on the request information with less conditional information, it is easy to cause the generated product promotion strategy to be inaccurate. For example, in the scenario where the target product is an installment payment service, the initial request information provided by the user is "What promotion strategy is there to encourage users to use installment payment?" In this request information, the known condition reflected only includes "installment payment", and it cannot reflect the characteristic data of the target user. Therefore, in order to improve the accuracy of the generated product promotion strategy, the characteristic data of the target user corresponding to the target product can be supplemented with the initial request information.

[0096] For example, for the request message "What promotion strategies are there to encourage users to pay in installments?", the supplemented request message generated after content supplementation processing can be "For the user group that pays in installments, the users are teenagers, their occupation is students, their consumption level is 0 to k yuan, the supply type information is personal e-commerce or shopping mall, the scenario information involving the target product is xxx online shopping, and the shopping items are clothing. What operational strategies are there to encourage users to pay in installments?" In the embodiments of this specification, the supplemented request message is only an example and does not constitute a specific limitation on the supplemented request message.

[0097] In the embodiments of this specification, the initial request information is supplemented with content based on the characteristic data of the product buyer and the product seller during the transaction involving the target product, so that the product promotion strategy generated for the request information after the content supplement can effectively affect the target users corresponding to the target product, for example, it can encourage the buyer to buy the target product or use the target product, and can improve the sales performance of the seller.

[0098] The characteristic data of the target users can be extracted from the historical behavior data of the target user group to improve the accuracy of the obtained characteristic data.

[0099] Optionally, obtaining the characteristic data of the target user corresponding to the target product may specifically include: performing intent analysis on the initial request information to predict the target user group corresponding to the target product; obtaining historical behavior data of the target user group; and determining the characteristic data based on the historical behavior data of the target user group.

[0100] In the embodiments of this specification, after receiving the initial request information provided by the user, the first agent will perform intent analysis on the request information. Intent analysis can refer to the process of analyzing the user's request information to identify the user's desired goal or behavior. For example, in response to the question "What promotional strategies are there to encourage users to use installment payments?", the intent analysis of this request information can be as follows: "This is an operational strategy question. The input is about encouraging users to use installment payments, but there are no user groups. I need to find the corresponding user groups first."

[0101] The first intelligent agent can predict the target user group corresponding to the target product based on the process of intent analysis of the initial request information. Specifically, the first intelligent agent can predict the target user group corresponding to the target product based on the large language model within the first intelligent agent, wherein the target user group may include at least one of the product buyer and the product seller in the transaction process involving the target product.

[0102] The target product can also be understood in an expanded way. For example, in addition to the target product itself, the target product can also include other similar products that are similar to the target product.

[0103] In the embodiments of this specification, obtaining historical behavior data of a target user group may include retrieving the target user group's historical behavior data from a database used to store system data through keyword retrieval, where the keyword may be identification information corresponding to the target user group, such as the target user group's title, occupation, or other identity information. The historical behavior data may refer to historical transaction information of the target user group during their shopping journey, which may include at least transaction date information, transaction amount information, information about the two parties to the transaction, details of the goods or services, payment method information, and transaction order information.

[0104] The characteristic data corresponding to the target user group can be determined from the acquired historical behavior data.

[0105] In the embodiments of this specification, the historical behavior data of the target user group is used to determine the characteristic data of the users that are supplemented in the initial request information, thereby generating a product promotion strategy for the prompt word information supplemented with the characteristic data of the target user group, so that the generated product promotion strategy can more effectively influence the target user group and improve the effectiveness of the product promotion strategy.

[0106] In an embodiment of the present specification, the first agent can perform intent analysis on the initial request information and analyze the user's characteristic data that needs to be supplemented in the initial request information, thereby avoiding the need to add other agents to the multi-agent system to complete the intent analysis task for the initial request information, thereby simplifying the structure of the hierarchical multi-agent system.

[0107] As an implementation method, a multi-agent system in a hierarchical model including a first agent and a second agent may also include a third agent for performing data analysis. The third agent can complete the intention analysis task for the initial request information, predict the target user group corresponding to the target product, and determine the characteristic data of the target user group, etc., thereby simplifying the work tasks of the first agent and reducing the computing resource usage of the first agent.

[0108] After extracting the prompt word information from the request information based on the function of the second agent, the correspondence between the prompt word information and the dimensions used to reflect the function of the second agent can be recorded, so that the second agent can quickly filter out the prompt word information that can be solved.

[0109] Optionally, after using the first intelligent agent to process the initial request information and obtain multiple prompt word information for multiple dimensions, it can also include: saving the first correspondence between each of the prompt word information and each of the dimensions; the dimension is the basis for processing the initial request information to obtain each of the prompt word information.

[0110] Correspondingly, the use of each second intelligent agent to generate a product segmentation promotion strategy corresponding to each prompt word information may specifically include: determining the dimension corresponding to each second intelligent agent based on the second correspondence between each second intelligent agent and each dimension used to reflect the function of the second intelligent agent; determining the prompt word information that each second intelligent agent can solve based on the dimension corresponding to each second intelligent agent and the first correspondence; using any second intelligent agent to solve the first prompt word information that any second intelligent agent can solve, and generating a product segmentation promotion strategy corresponding to the first prompt word information.

[0111] In an embodiment of the present specification, after the first agent extracts the corresponding designated prompt word information from the request information based on the function of each designated second agent, the first agent may save the corresponding relationship between the function of the designated second agent and the corresponding designated prompt word information. Since each function of the designated second agent may correspond to a designated dimension, the first corresponding relationship between the designated dimension corresponding to the function of the designated second agent and the designated prompt word information may also be saved. The designated dimension is the basis for extracting the designated prompt word information from the request information.

[0112] In the embodiments of this specification, the function of each second agent can be reflected through a dimension. For each second agent, a second corresponding relationship can exist between the second agent and the dimension used to reflect the function of the second agent. In actual applications, each second agent can also have multiple functions, that is, the function of a single agent can also be reflected through multiple dimensions, which is not limited to this.

[0113] In actual applications, for the second intelligent agent a, the method of determining the prompt word information that the second intelligent agent a can solve may include, based on the second correspondence relationship, under the condition that the second intelligent agent a is known, the dimension a corresponding to the second intelligent agent a can be determined; and then based on the first correspondence relationship, under the condition that the dimension a is known, the prompt word information a corresponding to the dimension a can be determined, thereby determining the prompt word information a as the prompt word information that the second intelligent agent a can solve.

[0114] After the second intelligent agent a determines the prompt word information a that can be solved, the second intelligent agent a can generate a product segmentation promotion strategy corresponding to the prompt word information a based on the large language model within the second intelligent agent a.

[0115] In actual applications, for each second agent, the method of determining the prompt word information that can be solved by the second agent a can be referred to to determine the prompt word information that can be solved by each second agent, which will not be repeated here.

[0116] In an embodiment of the present specification, after the prompt word information is disassembled from the request information, the first correspondence between each prompt word information and each dimension used to reflect the function of the second intelligent agent is saved, so that the prompt word information that each second intelligent agent can solve can be determined based on the first correspondence later, which is convenient and quick.

[0117] When the second intelligent agent solves the prompt word information that it can solve, it can also process it according to a certain workflow.

[0118] Optionally, the use of each second intelligent agent to generate a product segmentation promotion strategy corresponding to each prompt word information may specifically include: for any second intelligent agent, any second intelligent agent determines a second workflow for processing the prompt word information solved by any second intelligent agent, the second workflow is used to reflect the calling order of each functional tool required to be called by any second intelligent agent in the process of generating the product segmentation promotion strategy; based on the second workflow, the each functional tool is called in the calling order to generate a product segmentation promotion strategy for the prompt word information solved by any second intelligent agent.

[0119] In the embodiments of this specification, for each second intelligent agent, for example, a user operation intelligent agent used to provide strategies from the user dimension, after the user operation intelligent agent determines the prompt word information b that can be solved, the user operation intelligent agent can determine a second workflow for solving the prompt word information b. The second workflow can reflect the calling order of various functional tools required in the process of generating a product segmentation promotion strategy for the prompt word information b.

[0120] Among them, tools can refer to the specific technologies, frameworks, libraries or platforms used in the process of supporting large model work. These tools can provide more detailed and specific support for specific tasks or functions. They can be the specific execution units for realizing large model functions.

[0121] Based on the determined second workflow, the user operation agent can call the required tools to generate a product segmentation promotion strategy for prompt word information b. For example, the user operation agent can call the "Strategy Generation Tool" and generate a product segmentation promotion strategy for prompt word information b based on the "Strategy Generation Tool."

[0122] In an embodiment of the present specification, when the second intelligent agent solves the prompt word information that can be solved, it can generate a product segmentation promotion strategy based on the determined second workflow to improve the compliance of the second intelligent agent in solving the prompt word information, thereby improving the accuracy of the product segmentation promotion strategy generated by the second intelligent agent.

[0123] As an implementation method, the first agent may not supplement the content of the initial request information provided by the user, but directly decompose the initial request information into multiple prompt word information of multiple dimensions. After the second agent obtains the prompt word information that can be solved, the second agent may supplement the content of the prompt word information that can be solved. The content supplemented by the second agent for the prompt word information may also be the user's feature data. For the second agent with different functions, the content supplemented by the second agent for the prompt word information that can be solved may be the same or different. The second agent may supplement the prompt word information with content according to the function of the second agent, thereby improving the matching between the content supplemented for the prompt word information and the function of the second agent that solves the prompt word information, thereby improving the efficiency and accuracy of the second agent in solving the prompt word information that can be solved.

[0124] As another implementation method, the first intelligent agent disassembles and processes the request information after content supplementation to obtain multiple prompt word information in multiple dimensions. Before the second intelligent agent solves the prompt word information that can be solved, the second intelligent agent can also combine the functions of the second intelligent agent to supplement the content of the solved prompt word information again to assist the second intelligent agent in solving the prompt word information that can be solved, and improve the efficiency and accuracy of the second intelligent agent in solving the prompt word information that can be solved.

[0125] In order to improve the efficiency of the second intelligent agent in generating product segmentation promotion strategies, input information templates to be input into various functional tools can be prepared in advance.

[0126] Optionally, based on the second workflow, the various functional tools are called in the calling order to generate a product segmentation promotion strategy for the prompt word information solved by any of the second intelligent entities. Specifically, it may include: based on the prompt word information solved by any of the second intelligent entities, sending a first input parameter request information to a first interface for accessing the functional tool; the first input parameter request information is a request information generated based on a preset input parameter request template; obtaining first feedback information provided by the first interface; and generating a product segmentation promotion strategy for the prompt word information solved by any of the second intelligent entities based on the first feedback information.

[0127] In the embodiments of this specification, a functional tool can be understood as an entire system or a set of technical components used to implement a specific function. The first interface for accessing the functional tool can be an interface provided by the functional tool to external callers, which allows developers or clients to request the services provided by the functional tool in a predetermined format. The input parameter request information can represent all the input information that the user needs to provide when calling the interface. The input parameter request information defines the content, format and specific operations of the request. The preset input parameter request template can be a pre-established template for quickly generating input parameter request information. The first feedback information can be the response information corresponding to the first input parameter request information output by the first interface after the first input parameter request information is input into the first interface.

[0128] For example, let's assume the second agent is a user operation agent. The prompt information that the user operation agent can solve is: "For the user group that uses installment payment, the user is a teenager, the occupation is student, the consumption level is 0 to k yuan, the supply type information is personal e-commerce or shopping mall, the scenario information involving the target product is xxx online shopping, and the shopping item is clothing. What operational strategies are there from the user perspective to encourage users to use installment payment?"

[0129] The preset input parameter request template can be "market information: (), opportunity point direction: (), opportunity point name: (user's life stage (), consumption level (), occupation ())".

[0130] The first input parameter request information input into the first interface for accessing the "Strategy Output Tool" can be "market information: (xxx online shopping), opportunity point direction: (user operation), opportunity point name: (user's life stage (teenagers), consumption level (0 to k yuan), occupation (student))".

[0131] The first feedback information output by the first interface may be "We can provide an immediate discount of a yuan to young users, student users with a spending level between 0 and k yuan, which will cost b yuan in operating costs and is expected to generate c yuan in revenue."

[0132] In the embodiments of this specification, after obtaining the first feedback information output by the first interface, a product segmentation promotion strategy for the prompt word information resolved by the second agent can be generated based on the first feedback information. If the second agent needs to call multiple interfaces in the process of generating the product segmentation promotion strategy, if the first interface is the last interface called, the first feedback information output by the first interface can be the generated product segmentation promotion strategy. If the first interface is not the last interface called, the first feedback information can be used as input information to the next interface after the first interface, so as to generate the product segmentation promotion strategy based on the first feedback information.

[0133] In an embodiment of the present specification, in the process of calling the first interface for accessing the functional tool, the first input parameter request information input into the first interface can be generated based on a preset input parameter request template, thereby improving the efficiency of generating the first input parameter request information, and further improving the efficiency of generating the product promotion strategy.

[0134] If the prompt word information cannot provide all the information required by the preset input parameter request template, when the input parameter request information generated based on the preset input parameter request template is input into the first interface for accessing the functional tool, the first interface will not be able to output the first feedback information. Since the output information of the first interface may be the input information of other interfaces, if the first interface cannot output the first feedback information, it will cause the generation process of the product segmentation promotion strategy for the prompt word information to be stuck. In order to avoid the risk of the generation process being stuck, a second interface for accessing the functional tool can also be provided.

[0135] Optionally, after sending the first input parameter request information to the first interface for accessing the functional tool based on the prompt word information solved by any of the second intelligent entities, it may also include: if the first feedback information provided by the first interface is not obtained, sending the second input parameter request information to the second interface for accessing the functional tool based on the prompt word information solved by any of the second intelligent entities; the second input parameter request information is a request information generated based on the prompt word information solved by any of the second intelligent entities and the historical input parameter request information input into the second interface; generating a product segmentation promotion strategy for the prompt word information solved by any of the second intelligent entities based on the second feedback information provided by the second interface; the second interface is an interface with the same function as the first interface.

[0136] In an embodiment of the present specification, the second interface and the first interface can be interfaces for accessing the same functional tool. After calling the first interface for accessing the target functional tool, if the first interface does not output the corresponding first feedback information within the preset time, the second interface for accessing the target functional tool can be called. The second input parameter request information input to the second interface can be the input parameter request information customized by the large language model in the second intelligent agent after analyzing the solved prompt word information and the historical input parameter request information input to the second interface.

[0137] As an implementation method, the large language model in the second intelligent agent can extract the first feature data of the target user group from the solved prompt word information, and extract the second feature data related to the user from the historical input parameter request information input to the second interface, and generate the second input parameter request information based on the first feature data and the second feature data. Specifically, the second input parameter request information can be generated based on the same feature data between the first feature data and the second feature data. The method of generating the second input parameter request information proposed in the embodiment of this specification is only an example and does not constitute a limitation on the method of generating the second input parameter request information.

[0138] In the embodiment of this specification, the content of generating the product segmentation promotion strategy based on the second feedback information can refer to the content of generating the product segmentation promotion strategy based on the first feedback information, and will not be repeated here.

[0139] In the embodiments of this specification, the large language model in the second intelligent agent can also customize the input request information based on the resolved prompt word information and the historical input request information input into the second interface, so as to avoid the risk of jamming in the generation process of the product promotion strategy and improve the generation efficiency of the product promotion strategy.

[0140] In order to facilitate users to view product promotion strategies, the product promotion strategies generated for target products can be output in the form of preset templates.

[0141] Optionally, the product promotion strategy corresponding to the initial request information is obtained based on the product segmentation promotion strategy generated by each second intelligent agent, which may specifically include: obtaining a preset template corresponding to the initial request information for outputting the product promotion strategy; the preset template contains input areas corresponding to each dimension for filling in the product segmentation promotion strategy; according to the dimension, filling each product segmentation promotion strategy into each input area of ​​the preset template to obtain the product promotion strategy corresponding to the initial request information.

[0142] In an embodiment of the present specification, the preset template may include at least a first input area for inputting a summary of the response to the initial request information provided by the user, and a second input area for the user to input various product segmentation promotion strategies proposed for the target product in various dimensions.

[0143] For example, the preset template for outputting product promotion strategies is as follows:

[0144] Output preset templates for product promotion strategies.

[0145] Reply summary information: ().

[0146] The product segmentation promotion strategy proposed in the first dimension: ().

[0147] The product segmentation promotion strategy proposed in the second dimension: ().

[0148] The product segmentation promotion strategy proposed in the third dimension: ().

[0149] The product segmentation promotion strategy proposed in the fourth dimension: ().

[0150] It should be noted that the above preset templates are only examples and do not constitute a limitation on the specific form of the preset templates.

[0151] In the embodiment of this specification, the first intelligent agent will fill in the response summary information of the initial request information provided by the user, as well as the product segmentation promotion strategies proposed in each dimension, into the corresponding areas of the preset template.

[0152] For example, if the target product is an installment payment service, the initial request information for the target product may be "What promotion strategies are there to encourage users to pay in installments?" In response to this initial request information, in the user operation dimension, the product segmentation promotion strategy generated by the user operation agent is "You can provide an immediate discount of a yuan for teenage users and student users with a consumption level between 0 and k yuan. The estimated operating cost is b million and it is estimated to bring in c million in revenue." In the credit limit operation dimension, the product segmentation promotion strategy generated by the credit limit operation agent is "You can provide a payment limit increase for teenage users and student users with a consumption level between 0 and k yuan. The estimated operating cost is d million and it is estimated to bring in e million in revenue." In the product operation dimension, the product segmentation promotion strategy generated by the product operation agent is "It is observed that the preferred products for using the installment payment service are xxx category products, and there are no x-period transactions for this xxx category product. If x-period installment payments are supported for this xxx category product, it is estimated to bring in f million in revenue." In the supply operation dimension, the product segmentation promotion strategy generated by the supply operation agent is "No suitable strategy." Fill in the promotion strategies for each product segment into the preset template for outputting product promotion strategies. The product promotion strategies output to users are shown below.

[0153] Reply summary: (Regarding the question of what promotional strategies are there to encourage users to pay in installments, the following strategies are recommended from the four dimensions of user operation, credit limit operation, product operation, and supply operation).

[0154] The product segmentation promotion strategy proposed in the user operation dimension: (you can provide an immediate discount of a yuan for young users and student users with a consumption level of 0 to k yuan. The estimated operating cost is b million and it is estimated to bring in c million in revenue).

[0155] The product segmentation promotion strategy proposed in the credit limit operation dimension: (the payment limit can be increased for young users and student users with a consumption level of 0 to k yuan. It is estimated to cost d yuan in operating costs and is estimated to bring e yuan in revenue).

[0156] The product segmentation promotion strategy proposed in the product operation dimension: (It is observed that the preferred products using installment payment services are xxx category products, and there are no x-period transactions for this xxx category product. If x-period installment payments are supported for this xxx category product, it is estimated that it can bring f million in revenue).

[0157] Product segmentation promotion strategy proposed in the supply operation dimension: (no suitable strategy).

[0158] In the embodiment of this specification, the product promotion strategy is filled into a preset template and output to the user, thereby improving the convenience of the user in viewing the product promotion strategy.

[0159] Regarding the product segmentation promotion strategy generated by the second agent, before integrating the product segmentation promotion strategy into the product promotion strategy for the initial request information, it is necessary to determine the correlation between the product segmentation promotion strategy and the corresponding prompt word information.

[0160] Optionally, before obtaining the product promotion strategy corresponding to the initial request information based on each of the product segmentation promotion strategies generated by each of the second intelligent agents, the method may also include: for any of the product segmentation promotion strategies, judging whether the any product segmentation promotion strategy is correlated with the target prompt word information obtained by the second intelligent agent that generates the any product segmentation promotion strategy, and obtaining a first judgment result; if the first judgment result indicates that there is no correlation between the any product segmentation promotion strategy and the target prompt word information, updating the target prompt word information to obtain updated target prompt word information; the supplementary information relative to the initial request information contained in the updated target prompt word information is different from the supplementary information relative to the initial request information contained in the target prompt word information; or the updated target prompt word information includes several prompt word information obtained by decomposing the target prompt word information; and generating an updated product segmentation promotion strategy for the updated target prompt word information.

[0161] Correspondingly, the product promotion strategy corresponding to the initial request information is obtained based on each of the product segmentation promotion strategies generated by each of the second intelligent agents, which may specifically include: generating the product promotion strategy corresponding to the initial request information based on the updated product segmentation promotion strategy.

[0162] In the embodiments of the present specification, for each product segmentation promotion strategy generated by the second agent, for example, for the product segmentation promotion strategy a generated by the second agent a, the first agent determines whether there is a correlation between the product segmentation promotion strategy a and the prompt word information a solved by the second agent a. Specifically, the method of determining whether there is a correlation between the product segmentation promotion strategy a and the prompt word information a may include performing semantic analysis on the product segmentation promotion strategy a to extract its core meaning and obtain the corresponding semantic information a; then, performing feature extraction on the semantic information a to generate feature data a; next, performing semantic analysis on the prompt word information a to obtain its corresponding semantic information b; then, performing feature extraction on the semantic information b to obtain feature data b; calculating the similarity between the feature data a and the feature data b; if the similarity is less than a preset threshold, it means that there is no correlation between the product segmentation promotion strategy a and the prompt word information a; if the similarity is greater than or equal to the preset threshold, it means that there is a correlation between the two.

[0163] If it is determined that there is no correlation between the product segmentation promotion strategy a and the prompt word information a, the first agent may update the prompt word information a. The specific update method may include performing content supplementation processing on the prompt word information a. For example, based on the dimensional feature information carried in the prompt word information a, the feature data of the user used to assist in generating the strategy in terms of the dimensional feature is updated. The first supplementary information relative to the initial request information contained in the updated prompt word information a is different from the second supplementary information relative to the initial request information contained in the prompt word information a before the update, wherein the feature data in the first supplementary information may include all the feature data in the second supplementary information; or the feature data in the first supplementary information and the feature data in the second supplementary information are at least partially different.

[0164] The first agent may also update prompt word information a by performing a decomposition process on prompt word information a, decomposing prompt word information a into a plurality of prompt word information, for example, into prompt word information a1 and prompt word information a2. If the dimensional feature corresponding to prompt word information a is dimensional feature a, the dimensional features corresponding to prompt word information a1 and prompt word information a2 may be dimensional features obtained by subdividing dimensional feature a.

[0165] The method for the first agent to update prompt word information a may also include: performing content supplementation processing on prompt word information a to obtain content-supplemented prompt word information a; performing decomposition processing on the content-supplemented prompt word information a to decompose the content-supplemented prompt word information a into a plurality of prompt word information, such as prompt word information a3 and prompt word a4. If the dimensional feature corresponding to prompt word information a is dimensional feature a, the dimensional features corresponding to prompt word information a3 and prompt word information a4 may be dimensional features that are subdivided from dimensional feature a. The user feature data included in prompt word information a3 and the user feature data included in prompt word information a4 may be the same or different, and this is not limited to this.

[0166] If it is determined that there is no correlation between the product segmentation promotion strategy a and the prompt word information a, the first intelligent agent can update the prompt word information a to obtain the updated prompt word information a, and the second intelligent agent that solves the prompt word information a can solve the updated prompt word information a to generate an updated product segmentation promotion strategy a corresponding to the updated prompt word information a, and use the updated product segmentation promotion strategy a as the promotion strategy for the dimensional features corresponding to the prompt word information a.

[0167] In actual applications, if the updated prompt word information a includes prompt word information a5 and prompt word information a6, the second intelligent agent that solves the prompt word information a solves the prompt word information a5 and prompt word information a6, generates the product segmentation promotion strategy a5 corresponding to the prompt word information a5, and generates the product segmentation promotion strategy a6 corresponding to the prompt word information a6, and then summarizes the product segmentation promotion strategy a5 and the product segmentation promotion strategy a6 to obtain the promotion strategy for the dimensional features corresponding to the prompt word information a.

[0168] In the embodiment of this specification, only after determining that there is a correlation between the target product segmentation promotion strategy and the target prompt word information obtained by the second intelligent agent that generates the target product segmentation promotion strategy, will a product promotion strategy corresponding to the initial request information be generated based on the target product segmentation promotion strategy to improve the accuracy of the product promotion strategy.

[0169] In order to save the computing resources of the second agent, when the second agent generates a product segmentation promotion strategy in a certain dimension, it can limit the maximum number of times the product segmentation promotion strategy is generated in the dimension.

[0170] Optionally, before updating the target prompt word information, it may also include: if there is no correlation between any of the product segmentation promotion strategies and the target prompt word information, then determining whether the number of times the product segmentation promotion strategies are generated under the dimension corresponding to the target prompt word information is less than a preset number, and obtaining a second judgment result.

[0171] Correspondingly, updating the target prompt word information may specifically include:

[0172] If the second judgment result indicates that the number of times the product segmentation promotion strategy is generated under the dimension corresponding to the target prompt word information is less than a preset number, the target prompt word information is updated.

[0173] Correspondingly, the product promotion strategy corresponding to the initial request information is obtained based on the product segmentation promotion strategy generated by each second intelligent agent, which may specifically include: if the second judgment result indicates that the number of times the product segmentation promotion strategy is generated under the dimension corresponding to the target prompt word information is greater than or equal to the preset number, then the product promotion strategy corresponding to the initial request information is obtained based on the product segmentation promotion strategy corresponding to the target prompt word information.

[0174] In the embodiments of this specification, the preset number of times can be set according to actual needs, for example, 3, 5, etc., and this is not limited to this. Any product segmentation promotion strategy can be a promotion strategy generated for the target prompt word information. If any product segmentation promotion strategy is not related to the target prompt word information, the first agent must determine whether the number of times the product segmentation promotion strategy has been generated under the dimension corresponding to the target prompt word information is less than the preset number. The dimension corresponding to the target prompt word information can be understood as a dimension that reflects the function of the second agent capable of resolving the target prompt word information. If the number of times the product segmentation promotion strategy has been generated under the dimension corresponding to the target prompt word information is less than the preset number, the target prompt word information can be updated. If the number of times the product segmentation promotion strategy has been generated under the dimension corresponding to the target prompt word information is greater than or equal to the preset number, the target prompt word information may not be updated, and the product segmentation promotion strategy is used as the final product segmentation promotion strategy generated for the target prompt word information. Based on the final product segmentation promotion strategy generated for the target prompt word information, a product promotion strategy for the initial request information is obtained.

[0175] In an embodiment of the present specification, if the number of times the second agent generates product segmentation promotion strategies for the initial request information in a dimension used to reflect the function of the second agent reaches a preset number, the product segmentation promotion strategy most recently generated by the second agent in this dimension can be used as the final product segmentation promotion strategy in this dimension to reduce excessive consumption of computing resources of the second agent.

[0176] As an implementation method, if the product segmentation promotion strategy generated by the second intelligent agent for the initial request information in the dimension used to reflect the function of the second intelligent agent includes multiple first product segmentation promotion strategies, then each first product segmentation promotion strategy can be matched with the first prompt word information split out in this dimension, and the similarity between each first product segmentation promotion strategy and the first prompt word information can be determined, and the first product segmentation promotion strategy corresponding to the maximum similarity can be used as the final product segmentation promotion strategy in this dimension.

[0177] In the embodiments of this specification, the multi-agent system based on the hierarchical model composed of the first agent and the second agent may also include a fourth agent. The fourth agent may be an agent for automatically deploying the product promotion strategy output by the first agent. The fourth agent may deploy the generated product promotion strategy on the product seller in the transaction process involving the target product, so as to influence the transaction behavior of the product buyer, for example, influencing the payment method of the product buyer.

[0178] In the embodiments of this specification, for a target product, discovering the opportunity points for increasing the target product's revenue, generating corresponding product promotion strategies for the opportunity points, and deploying the generated product promotion strategies can all be fully automatically managed by a hierarchical multi-agent system, thereby simplifying the work tasks of operators and reducing their human resource cost expenditures.

[0179] In order to more clearly illustrate the product promotion strategy generation method provided in this manual, Figure 4 This is a swim lane diagram of a method for generating a product promotion strategy provided by an embodiment of this specification. Figure 4 As shown, it includes steps 402 to 418.

[0180] Step 402: The first agent obtains the request provided by the user and generates initial request information for a promotion strategy for a target product.

[0181] The target product may be an installment payment service, and the corresponding initial request information may be "What promotion strategies are there to encourage users to use installment payment?"

[0182] Step 404: The first agent performs content supplementation processing on the initial request information to obtain request information after content supplementation.

[0183] The content supplemented with respect to the initial request information may be feature data of the target user corresponding to the target product.

[0184] Step 406: The first agent disassembles and processes the request information after content supplementation to obtain multiple prompt word information in multiple dimensions.

[0185] Among them, the first intelligent agent can disassemble and process the request information after content supplementation according to the function of the second intelligent agent, and disassemble the request information after content supplementation into prompt word information corresponding to each dimension according to each dimension used to reflect the function of the second intelligent agent.

[0186] Step 408: Each second agent obtains the prompt word information that it can solve.

[0187] Step 410: Each second agent solves the prompt word information that it can solve, and obtains the product segmentation promotion strategy corresponding to each prompt word information.

[0188] Step 412: The first agent determines, for each product segmentation promotion strategy, whether there is a correlation between the product segmentation promotion strategy and the target prompt word information solved by the second agent that generates the product segmentation promotion strategy.

[0189] Step 414: If there is a correlation between any of the product segmentation promotion strategies and the target prompt word information, the first agent generates a product promotion strategy for the initial request information based on any of the product segmentation promotion strategies.

[0190] Step 416: If there is no correlation between any of the product segmentation promotion strategies and the target prompt word information, the first agent updates the target prompt word information to obtain updated target prompt word information.

[0191] Step 418: The first agent generates a product promotion strategy for the initial request information based on the product segmentation promotion strategy generated by the second agent for the updated target prompt word information.

[0192] Based on the same idea, the embodiments of this specification also provide a device corresponding to the above method. Figure 5 This specification provides the corresponding embodiment Figure 2 A structural diagram of a device for generating a product promotion strategy. Figure 5 As shown, the device may include:

[0193] An initial request information acquisition module 502 is used to acquire initial request information provided by a user to generate a promotion strategy for a target product;

[0194] Initial request information processing module 504 is configured to process the initial request information using the first agent to obtain multiple prompt word information for multiple dimensions; different prompt word information is used to describe the request for generating a promotion strategy for the target product from different dimensions; the multiple dimensions include at least one of a user dimension and a product dimension;

[0195] A product segmentation promotion strategy generation module 506 is configured to generate a product segmentation promotion strategy corresponding to each of the prompt word information using each of the second agents;

[0196] The product promotion strategy obtaining module 508 is used to obtain the product promotion strategy corresponding to the initial request information based on each of the product segmentation promotion strategies generated by each of the second intelligent agents.

[0197] based on Figure 5 The present specification also provides some specific implementation plans, which are described below.

[0198] Optionally, the device may further include:

[0199] The first workflow determination module is used to determine a first workflow for processing the initial request information using the first agent in response to the initial request information, wherein the first workflow is used to reflect the processing logic for processing the initial request information.

[0200] The second processing module for the initial request information is used to process the initial request information using the first agent based on the first workflow.

[0201] Optionally, the initial request information processing module 504 may specifically include:

[0202] The content supplement processing unit is used to perform content supplement processing on the initial request information to obtain request information after supplement processing.

[0203] The disassembly processing unit is used to disassemble the request information after the supplementary processing based on the functions of each second intelligent agent to obtain multiple prompt word information of the multiple dimensions; the prompt word information of one dimension at least meets the function of one second intelligent agent.

[0204] Optionally, the content supplement processing unit may specifically include:

[0205] The feature data acquisition subunit is used to acquire the feature data of the target user corresponding to the target product.

[0206] The content supplement processing subunit is used to perform content supplement processing on the initial request information based on the feature data to obtain request information after supplement processing; the request information after supplement processing includes the feature data.

[0207] Optionally, the feature data acquisition subunit may be specifically used to:

[0208] An intent analysis is performed on the initial request information to predict the target user group corresponding to the target product.

[0209] Obtain historical behavior data of the target user group.

[0210] The characteristic data is determined based on the historical behavior data of the target user group.

[0211] Optionally, the device may further include:

[0212] The first correspondence storage module is configured to store a first correspondence between each prompt word information and each dimension; the dimension is a basis for processing the initial request information to obtain each prompt word information.

[0213] The product segmentation promotion strategy generation module 506 may specifically include:

[0214] The dimension determination unit is used to determine the dimension corresponding to each second agent based on the second corresponding relationship between each second agent and each dimension used to reflect the function of the second agent.

[0215] The prompt word information determining unit is used to determine the prompt word information that each second agent can solve based on the dimension corresponding to each second agent and the first corresponding relationship.

[0216] The product segmentation promotion strategy generating unit is configured to utilize any of the second intelligent agents to solve the first prompt word information that can be solved by any of the second intelligent agents, and generate a product segmentation promotion strategy corresponding to the first prompt word information.

[0217] Optionally, the product segmentation promotion strategy generating module 506 may specifically include:

[0218] The second workflow determination unit is used to determine, for any second intelligent agent, a second workflow for processing the prompt word information solved by any second intelligent agent, and the second workflow is used to reflect the calling order of various functional tools that need to be called by any second intelligent agent in the process of generating the product segmentation promotion strategy.

[0219] A function tool calling unit is used to call each function tool in the calling order based on the second workflow to generate a product segmentation promotion strategy for the prompt word information solved by any of the second intelligent agents.

[0220] Optionally, the function tool calling unit may specifically include:

[0221] a first input parameter request information sending subunit, configured to send a first input parameter request information to a first interface for accessing the functional tool based on the prompt word information solved by any of the second agents; the first input parameter request information being request information generated based on a preset input parameter request template;

[0222] The first feedback information obtaining subunit is configured to obtain first feedback information provided by the first interface.

[0223] The first product segmentation promotion strategy generating sub-unit is used to generate a product segmentation promotion strategy for the prompt word information solved by any of the second intelligent agents based on the first feedback information.

[0224] Optionally, the function tool calling unit may further include:

[0225] The second input parameter request information sending sub-unit is used to send the second input parameter request information to the second interface for accessing the functional tool based on the prompt word information solved by any of the second intelligent agents if the first feedback information provided by the first interface is not obtained; the second input parameter request information is generated based on the prompt word information solved by any of the second intelligent agents and the historical input parameter request information input into the second interface.

[0226] The second generation sub-unit of the product segmentation promotion strategy is used to generate a product segmentation promotion strategy for the prompt word information solved by any of the second intelligent agents based on the second feedback information provided by the second interface; the second interface is an interface with the same function as the first interface.

[0227] Optionally, the product promotion strategy obtaining module 508 may specifically include:

[0228] The preset template acquisition unit is used to obtain a preset template corresponding to the initial request information for outputting the product promotion strategy; the preset template includes various input areas corresponding to various dimensions for filling in the product segmentation promotion strategy.

[0229] The product segmentation promotion strategy filling unit is used to fill each of the product segmentation promotion strategies into each of the input areas of the preset template according to the dimensions, so as to obtain the product promotion strategy corresponding to the initial request information.

[0230] Optionally, the device may further include:

[0231] The first judgment module is used to judge whether there is a correlation between any product segmentation promotion strategy in each of the product segmentation promotion strategies and the target prompt word information obtained by the second intelligent agent that generates the any product segmentation promotion strategy, and obtain a first judgment result.

[0232] a target prompt word information updating module configured to update the target prompt word information to obtain updated target prompt word information if the first judgment result indicates that there is no correlation between any of the product segmentation promotion strategies and the target prompt word information; wherein the supplementary information relative to the initial request information included in the updated target prompt word information is different from the supplementary information relative to the initial request information included in the target prompt word information; or wherein the updated target prompt word information includes multiple pieces of prompt word information obtained by decomposing the target prompt word information.

[0233] The updated product segmentation promotion strategy generating module is used to generate an updated product segmentation promotion strategy for the updated target prompt word information.

[0234] The product promotion strategy obtaining module 508 may specifically include:

[0235] A product promotion strategy generating unit is configured to generate a product promotion strategy corresponding to the initial request information based on the updated product segmentation promotion strategy.

[0236] Optionally, the device may further include:

[0237] The second judgment module is used to judge whether the number of times the product segmentation promotion strategy is generated under the dimension corresponding to the target prompt word information is less than a preset number if there is no correlation between any of the product segmentation promotion strategies and the target prompt word information, and obtain a second judgment result.

[0238] The target prompt word information updating module may specifically include:

[0239] The target prompt word information updating unit is configured to update the target prompt word information if the second judgment result indicates that the number of times the product segmentation promotion strategy is generated under the dimension corresponding to the target prompt word information is less than a preset number.

[0240] The product promotion strategy obtaining module 508 may specifically include:

[0241] A product promotion strategy obtaining unit is used to obtain the product promotion strategy corresponding to the initial request information based on the product promotion strategy corresponding to the target prompt word information if the second judgment result indicates that the number of times the product segmentation promotion strategy is generated under the dimension corresponding to the target prompt word information is greater than or equal to the preset number.

[0242] Based on the same idea, the embodiments of this specification also provide devices corresponding to the above methods.

[0243] Figure 6 This is a schematic diagram of the structure of a product promotion strategy generation device provided in the embodiment of this specification. Figure 6 As shown, the device 600 may include:

[0244] at least one processor 610; and,

[0245] A memory 630 is communicatively coupled to the at least one processor.

[0246] Among them, corresponding to Figure 2In the method shown, the memory 630 stores instructions 620 that can be executed by the at least one processor 610, and the instructions are executed by the at least one processor 610 to enable the at least one processor 610 to: obtain the request provided by the user to generate initial request information for a promotion strategy for a target product; use the first intelligent agent to process the initial request information to obtain multiple prompt word information for multiple dimensions; different prompt word information is used to describe the request for generating a promotion strategy for the target product from different dimensions; the multiple dimensions include at least one dimension in the user dimension and the product dimension; use each second intelligent agent to generate a product segmentation promotion strategy corresponding to each prompt word information; based on each product segmentation promotion strategy generated by each second intelligent agent, obtain the product promotion strategy corresponding to the initial request information.

[0247] Based on the same idea, the embodiments of this specification also provide a computer-readable medium corresponding to the above method. The computer-readable medium stores computer-readable instructions, which can be executed by a processor to implement the above product promotion strategy generation method.

[0248] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments. Figure 6 As for the device shown, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0249] 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 regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0250] In the 1990s, technological improvements could be clearly distinguished as either hardware improvements (for example, improvements to circuit structures like diodes, transistors, and switches) or software improvements (improvements to process flows). However, with the advancement of technology, many process flow improvements today can now be considered direct improvements to hardware circuit structures. Designers almost always create the corresponding hardware circuit structure by programming the improved process flow into the hardware circuit. Therefore, it cannot be said that a process flow improvement cannot be implemented using hardware modules. For example, a programmable logic device (PLD), such as a field programmable gate array (FPGA), is an integrated circuit whose logical function is determined by user programming. Designers can "integrate" a digital system on a PLD through their own programming, without having to hire a chip manufacturer to design and manufacture a dedicated integrated circuit chip. Moreover, nowadays, instead of manually fabricating integrated circuit chips, this programming is mostly done using "logic compiler" software. This is similar to the software compiler used when developing programs. Before compilation, the original code must also be written in a specific programming language, called a hardware description language (HDL). There is not just one HDL, but many, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, RHDL (Ruby Hardware Description Language), etc. The most commonly used ones are VHDL (Very-High-Speed ​​Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art will also understand that by simply programming the method flow in one of these hardware description languages ​​and then programming it into an integrated circuit, a hardware circuit that implements the logic method flow can be easily obtained.

[0251] The controller can be implemented in any suitable manner. For example, the controller can take the form of a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro)processor, logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers, and embedded microcontrollers. Examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051 F320. The memory controller can also be implemented as part of the control logic of the memory. Those skilled in the art will also know that in addition to implementing the controller in a purely computer-readable program code format, it is entirely possible to implement the same functionality in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, such a controller can be considered a hardware component, and the devices included therein for implementing various functions can also be considered as structures within the hardware component. Or even, the devices for implementing various functions can be considered as both software modules that implement the method and structures within the hardware component.

[0252] The systems, devices, modules, or units described in the above embodiments may be implemented by computer chips or entities, or by products having certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0253] For the convenience of description, the above devices are described as being divided into various units according to their functions. Of course, when implementing this application, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0254] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0255] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, 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 produce 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 flowcharts and / or block diagrams. 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.

[0256] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work 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 The function specified in one or more boxes.

[0257] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0258] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0259] 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.

[0260] 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. The 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 memory (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, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic 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. As defined herein, computer-readable media does not include transitory media such as modulated data signals and carrier waves.

[0261] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0262] 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 take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0263] The present application may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communications network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0264] The foregoing is merely an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should all be included within the scope of the claims of the present application.

Claims

1. A method for generating a product promotion strategy, the method being applied to a product promotion strategy generation system. The system comprises a first agent for generating multiple prompt word information in multiple dimensions based on a user's initial request information, and a second agent for generating a product promotion strategy for each dimension of prompt word information. The method comprises: Obtaining the request provided by the user to generate initial request information for a promotion strategy for a target product; The first agent processes the initial request information to obtain a plurality of prompt word information for a plurality of dimensions; different prompt word information is used to describe the request for generating a promotion strategy for the target product from different dimensions; the plurality of dimensions include at least one of a user dimension and a product dimension; Utilizing each of the second intelligent agents, generating a product segmentation promotion strategy corresponding to each of the prompt word information; Based on each of the product segmentation promotion strategies generated by each of the second intelligent agents, the product promotion strategy corresponding to the initial request information is obtained.

2. The method according to claim 1, before using the first agent to process the initial request information to obtain multiple prompt word information for multiple dimensions, further comprising: In response to the initial request information, using the first agent to determine a first workflow for processing the initial request information, wherein the first workflow is used to reflect a processing logic for processing the initial request information; Based on the first workflow, the first agent is used to process the initial request information.

3. The method according to claim 1, wherein the first agent processes the initial request information to obtain multiple prompt word information for multiple dimensions, specifically comprising: Performing content supplementation processing on the initial request information to obtain supplemented request information; Based on the functions of each of the second agents, the request information after the supplementary processing is decomposed to obtain the plurality of prompt word information of the plurality of dimensions; The prompt word information of one dimension conforms to at least one function of the second intelligent agent.

4. The method according to claim 3, wherein the performing content supplementation processing on the initial request information to obtain the supplemented request information specifically comprises: Obtaining characteristic data of target users corresponding to the target product; Based on the characteristic data, content supplementation processing is performed on the initial request information to obtain supplemented request information; the supplemented request information includes the characteristic data.

5. The method according to claim 4, wherein obtaining characteristic data of a target user corresponding to the target product specifically comprises: Performing intent analysis on the initial request information to predict the target user group corresponding to the target product; Obtaining historical behavior data of the target user group; The characteristic data is determined based on the historical behavior data of the target user group.

6. The method according to claim 1, further comprising: Saving the first correspondence between each of the prompt word information and each of the dimensions; The dimension is a basis for processing the initial request information to obtain the respective prompt word information; The step of utilizing each of the second agents to generate a product segmentation promotion strategy corresponding to each of the prompt word information specifically includes: determining the dimensions corresponding to the respective second agents according to the second corresponding relationships between the respective second agents and the respective dimensions used to embody the functions of the second agents; Determining prompt word information that each second agent can solve based on the dimensions corresponding to each second agent and the first corresponding relationship; Any of the second intelligent agents is used to solve the first prompt word information that can be solved by any of the second intelligent agents, and a product segmentation promotion strategy corresponding to the first prompt word information is generated.

7. The method according to claim 1, wherein the generating of product segmentation promotion strategies corresponding to each of the prompt word information by using each of the second agents specifically comprises: For any of the second agents, the second agent determines a second workflow for processing the prompt word information solved by the second agent, where the second workflow reflects the order in which the functional tools are called when the second agent generates the product segmentation promotion strategy. Based on the second workflow, the various functional tools are called in the calling order to generate a product segmentation promotion strategy for the prompt word information solved by any of the second intelligent agents.

8. The method according to claim 7, wherein the method, based on the second workflow, calls the functional tools in the calling order to generate a product segmentation promotion strategy for the prompt word information solved by any of the second agents, specifically comprises: Based on the prompt word information solved by any of the second agents, sending first input request information to a first interface for accessing the functional tool; The first input parameter request information is request information generated based on a preset input parameter request template; Obtaining first feedback information provided by the first interface; Based on the first feedback information, a product segmentation promotion strategy is generated for the prompt word information solved by any of the second intelligent agents.

9. The method according to claim 8, further comprising: after sending a first input parameter request message to a first interface for accessing the function tool based on the prompt word information solved by any of the second agents; If the first feedback information provided by the first interface is not obtained, then based on the prompt word information solved by any of the second agents, a second input parameter request information is sent to the second interface for accessing the functional tool; the second input parameter request information is generated based on the prompt word information solved by any of the second agents and the historical input parameter request information input to the second interface; generating, based on the second feedback information provided by the second interface, a product segmentation promotion strategy for the prompt word information solved by any of the second agents; The second interface is an interface having the same function as the first interface.

10. The method according to claim 1, wherein obtaining the product promotion strategy corresponding to the initial request information based on the product segmentation promotion strategy generated by each second agent specifically comprises: Obtaining a preset template corresponding to the initial request information for outputting the product promotion strategy; the preset template includes input areas corresponding to various dimensions for filling in the product segmentation promotion strategy; According to the dimensions, each of the product segmentation promotion strategies is filled into each of the input areas of the preset template to obtain a product promotion strategy corresponding to the initial request information.

11. The method according to claim 1, wherein before obtaining the product promotion strategy corresponding to the initial request information based on the product segmentation promotion strategy generated by each second agent, the method further comprises: For each of the product segmentation promotion strategies, determining whether there is a correlation between the product segmentation promotion strategy and the target prompt word information obtained by the second agent that generates the product segmentation promotion strategy, thereby obtaining a first determination result; If the first judgment result indicates that there is no correlation between any of the product segmentation promotion strategies and the target prompt word information, then updating the target prompt word information to obtain updated target prompt word information; The supplementary information relative to the initial request information included in the updated target prompt word information is different from the supplementary information relative to the initial request information included in the target prompt word information; or the updated target prompt word information includes multiple prompt word information obtained by decomposing the target prompt word information; generating an updated product segmentation promotion strategy for the updated target prompt word information; The obtaining of the product promotion strategy corresponding to the initial request information based on the product segmentation promotion strategy generated by each second agent specifically includes: Based on the updated product segmentation promotion strategy, a product promotion strategy corresponding to the initial request information is generated.

12. The method according to claim 11, before updating the target prompt word information, further comprising: If there is no correlation between any of the product segmentation promotion strategies and the target prompt word information, determining whether the number of times the product segmentation promotion strategies are generated under the dimension corresponding to the target prompt word information is less than a preset number, and obtaining a second determination result; The target prompt word information is updated, specifically including: If the second judgment result indicates that the number of times the product segmentation promotion strategy is generated under the dimension corresponding to the target prompt word information is less than a preset number, then updating the target prompt word information; The obtaining of the product promotion strategy corresponding to the initial request information based on the product segmentation promotion strategy generated by each second agent specifically includes: If the second judgment result indicates that the number of times the product segmentation promotion strategy is generated under the dimension corresponding to the target prompt word information is greater than or equal to the preset number, then the product promotion strategy corresponding to the initial request information is obtained based on the product segmentation promotion strategy corresponding to the target prompt word information.

13. A device for generating a product promotion strategy, comprising a product promotion strategy generation system, the product promotion strategy generation system comprising a first agent for generating multiple prompt word information in multiple dimensions based on a user's initial request information, and a second agent for generating a product promotion strategy for each dimension of prompt word information. The device comprises: An initial request information acquisition module is used to obtain initial request information provided by a user to generate a promotion strategy for a target product; an initial request information processing module, configured to process the initial request information using the first agent to obtain a plurality of prompt word information for a plurality of dimensions; different prompt word information is used to describe a request for generating a promotion strategy for the target product from different dimensions; the plurality of dimensions including at least one of a user dimension and a product dimension; A product segmentation promotion strategy generation module, configured to generate a product segmentation promotion strategy corresponding to each of the prompt word information using each of the second agents; The product promotion strategy obtaining module is used to obtain the product promotion strategy corresponding to the initial request information based on each of the product segmentation promotion strategies generated by each of the second intelligent agents.

14. A device for generating a product promotion strategy, comprising: at least one processor; as well as, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor. The instructions are executed by the at least one processor to enable the at least one processor to execute the product promotion strategy generating method according to any one of claims 1 to 12.

15. A computer-readable medium having computer-readable instructions stored thereon, wherein the computer-readable instructions can be executed by a processor to implement the product promotion strategy generating method according to any one of claims 1 to 12.

Images