Resource information recommendation methods, devices, storage media and program products

Abstract

This application provides a resource information recommendation method, device, storage medium, and program product. In this application embodiment, an AI resource information recommendation model is used to determine the first target resource configuration information from various resource configuration information based on the recommendation conditions of a first user's potential multi-dimensional resource needs and resource configuration information. This allows for rapid and accurate matching of suitable resource configuration information to the user. Furthermore, the language style information of the first user can be analyzed based on the first user's initial and in-depth profile data. Based on this, recommendation reason information and script information for recommending the first target resource configuration information to the first user can be generated. Target service personnel can use the script techniques provided by the script information to recommend the first target resource configuration information to the first user, improving the persuasiveness of the recommended resource configuration information. Furthermore, by adapting the language style to the first user, the user's acceptance of the recommendation process is increased, thereby improving the conversion rate of resource configuration information.

Description

Technical Field

[0001] This application relates to the field of computer processing technology, and in particular to a method, apparatus, storage medium, and program product for recommending resource information. Background Technology

[0002] With the development of the internet industry, internet services are becoming increasingly diverse to meet the needs of different customers. Taking platform services as an example, these platforms offer a variety of services, each with varying customer benefits. Currently, faced with numerous service options, platform sales personnel can rely on experience and intuition to recommend suitable services to customers. However, this traditional recommendation method lacks data support and personalized recommendation capabilities, making it difficult to quickly and accurately recommend the most suitable service to users. Furthermore, the sales personnel's recommendation scripts often depend on their own experience, lacking persuasive power, thus impacting the platform service conversion rate. Summary of the Invention

[0003] This application provides a resource information recommendation method, device, storage medium, and program product to quickly and accurately match suitable resource configuration information for users, and to generate recommendation reason information and persuasive information that are compatible with the recommended resource configuration information. This allows the persuasiveness of the recommended resource configuration information to be improved by using persuasive techniques provided by the persuasive information and recommendation reason information to recommend the resource configuration information to users, thereby increasing the conversion rate of the recommended resource configuration information.

[0004] This application provides a resource information recommendation method applied to a recommendation platform. The method includes: responding to a resource information recommendation request submitted by a target service provider in a target service domain; obtaining initial profile data and deep profile data of a first user, wherein the initial profile data is obtained through user profile analysis based on the first user's user information, and the deep profile data is obtained through user profile analysis based on the conversation content between the target service provider and the first user; performing potential resource demand analysis on the first user based on the initial profile data and deep profile data to obtain potential multi-dimensional resource demand information of the first user; and inputting the initial profile data, deep profile data, multi-dimensional resource demand information, resource configuration information, and their recommendation conditions into an AI resource information recommendation model, based on the multi-dimensional resource demand information and... The system recommends resource allocation information based on various criteria, selecting a first target resource allocation information from multiple sources and generating a recommendation reason for the first target resource allocation information. It analyzes the language style information of the first user based on initial and deep user profile data, and generates a script to recommend the first target resource allocation information to the first user using the recommendation reason information. The AI ​​resource information recommendation model is fine-tuned from a pre-trained model based on sample data from the target service domain. The first target resource allocation information, recommendation reason information, and script information are sent to the first terminal of the target service personnel, allowing them to recommend the first target resource allocation information to the first user using the script information and recommendation reason information through the communication channel between the first terminal and the first user's second terminal.

[0005] This application also provides an electronic device, including: a memory and a processor; the memory for storing a computer program; and the processor, coupled to the memory, for executing the computer program to implement the steps in the above method.

[0006] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps in the above-described method.

[0007] This application also provides a computer program product, which includes a computer program / instructions that, when executed by a processor, enable the processor to perform the steps in the above-described method.

[0008] In this embodiment, an AI resource information recommendation model is used to determine the first target resource configuration information from various resource configuration information based on the first user's potential multi-dimensional resource needs and resource configuration information. This allows for quick and accurate matching of suitable resource configuration information to the user. Furthermore, the model can analyze the first user's language style information based on initial and in-depth profile data. Based on this, it generates recommendation reasons and scripts for recommending the first target resource configuration information to the first user. Service personnel can then use the scripts provided to recommend the first target resource configuration information to the first user, improving the persuasiveness of the recommended resource configuration information. The model also adapts the language style to the first user, increasing user acceptance of the recommendation process and improving the conversion rate of the resource configuration information. Attached Figure Description

[0009] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0010] Figure 1 A schematic diagram of the structure of a resource information recommendation system provided in an exemplary embodiment of this application;

[0011] Figure 2a A flowchart illustrating a resource information recommendation method provided for an exemplary embodiment of this application;

[0012] Figure 2b A flowchart illustrating another resource information recommendation method provided as another exemplary embodiment of this application;

[0013] Figure 2c A flowchart illustrating yet another resource information recommendation method provided as an exemplary embodiment of this application;

[0014] Figure 2d A schematic flowchart illustrating the model training process provided for an exemplary embodiment of this application;

[0015] Figure 2e A flowchart illustrating the resource information generation process provided for an exemplary embodiment of this application;

[0016] Figure 3 A schematic diagram of the structure of an electronic device provided for an exemplary embodiment of this application. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0018] 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, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0019] Users can enjoy corresponding user benefits by purchasing resource allocation information. Users can be B-end enterprises providing services or goods, or C-end users seeking services or goods. Taking posting as an example, a B-end enterprise needs to promote its services or goods through posting, so it needs posting privileges to increase the post's exposure. Therefore, it needs to purchase resource allocation information to enjoy the corresponding posting privileges. Similarly, a C-end user needs to purchase services or goods but lacks purchasing experience and needs the platform to recommend services or goods that match their needs. In this case, they can purchase resource allocation information to enjoy the corresponding recommendation privileges. Or, a user can post seeking services or goods, so they need posting privileges to increase the post's exposure. Therefore, they need to purchase resource allocation information to enjoy the corresponding posting privileges.

[0020] Currently, faced with a multitude of resource allocation options, platform sales personnel can rely on experience and intuition to recommend suitable resource allocation information to customers. However, this traditional recommendation method lacks data support and personalized recommendation capabilities, making it difficult to quickly and accurately recommend the most suitable resource allocation information to users. Moreover, the sales personnel's recommendation scripts often depend on their own experience accumulation, resulting in weak persuasiveness and thus affecting the conversion rate of platform resource allocation information.

[0021] To address the aforementioned technical issues, this application embodiment utilizes an AI resource information recommendation model. Based on the recommendation conditions of the first user's potential multi-dimensional resource needs and resource configuration information, it determines the first target resource configuration information from various resource configuration information sources, quickly and accurately matching suitable resource configuration information to the user. Furthermore, it can analyze the first user's language style information based on the first user's initial and in-depth profile data, thereby generating recommendation reasons and scripts for recommending the first target resource configuration information to the first user. Target service personnel can use the scripts provided to recommend the first target resource configuration information to the first user based on the recommendation reasons, improving the persuasiveness of the recommended resource configuration information. It also adapts the language style to the first user, increasing user acceptance of the recommendation process and improving the conversion rate of the resource configuration information.

[0022] The technical solutions provided in the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0023] Figure 1 This is a schematic diagram of the structure of a resource information recommendation system provided for an exemplary embodiment of this application. For example... Figure 1 As shown, the resource information recommendation system includes a recommendation platform 101 and a first terminal 102, which can communicate with each other. It should be noted that... Figure 1 This is merely an illustrative diagram and does not constitute a limitation on the technical solution of this application.

[0024] In this embodiment, the first terminal is the terminal device used by the target service personnel in the target field. This embodiment does not limit the implementation form of the first terminal. For example, the first terminal can be a smartphone, tablet, laptop, or desktop computer, etc.; another example is that the first terminal can also be a smart wearable device, such as a smartwatch, smart bracelet, etc.; yet another example is that the first terminal can also be various smart home appliances with display screens, such as smart TVs, smart large screens, or smart robots, etc. Furthermore, various software applications can be installed on the first terminal. These software applications can be standalone apps, mini-programs that depend on apps, or web pages; this embodiment does not limit this. Software applications installed on the first terminal include, but are not limited to: instant messaging applications and service applications. Instant messaging applications refer to applications that allow two or more people to exchange messages in real time over a network. They support users sending text messages, pictures, videos, voice messages, file transfers, etc., and usually provide voice and video call functions. Instant messaging applications can be, for example, conversational tools such as WeChat and Weibo. Service applications are used to provide services or goods. Examples of service applications include recruitment applications, rental applications, shopping applications, ride-hailing applications, and other applications providing specialized services, or comprehensive service applications that integrate various services and goods. This embodiment does not limit the scope of such applications. Service applications may also include conversational functions, including but not limited to customer service consultation services and call services. Customer service consultation services refer to consultation services provided by the application's customer service representatives to users, and these consultation services can be conducted through a conversational interface. Call services refer to the application's function of allowing users to make calls to companies providing services or goods. Furthermore, when the recommendation platform is implemented as a software application, it can be installed on the primary terminal of the target service personnel, or on other terminals besides those used by the service personnel. This embodiment does not limit the scope of such installations.

[0025] The recommendation platform can store various resource configuration information for the target domain and user information for multiple users. These users refer to registered users of other applications associated with the recommendation platform, such as users of service applications. Users can be B-end enterprises providing services or goods, or C-end users demanding services or goods. In an optional embodiment, the recommendation platform can be implemented as a software application. This software application can have high processing power and be capable of performing tasks such as... Figure 2aThe steps of the relevant embodiments achieve the goal of sending the resource configuration information matched for the user, its configuration content, recommendation reason information, and script information to the first terminal of the target service personnel, so that the target service personnel can use the script skills provided by the script information to recommend the resource configuration information to the user based on the recommendation reason information. This embodiment does not limit the specific implementation form of the software application; the software application can be a standalone APP, a mini-program that depends on an APP, or a webpage. In another optional embodiment, the recommendation platform can also be implemented as a server, which can perform actions such as... Figure 2a The steps of the relevant embodiments are described below. This embodiment does not limit the specific implementation form of the server-side of the recommendation platform; the server-side can be a single physical server, a cloud server, or a server array.

[0026] Optionally, the resource information recommendation system further includes a second terminal 103. The second terminal can be a terminal device used by a B-end enterprise or a C-end user; for ease of description and distinction, it is referred to as the first user. This embodiment does not limit the specific implementation of the second terminal; please refer to the relevant description of the first terminal, which will not be repeated here. The recommendation platform sends the resource configuration information matched for the first user, its configuration content, recommendation reason information, and script information to the first terminal of the target service personnel. This allows the target service personnel to recommend resource configuration information to the first user using the script techniques provided by the script information and the recommendation reason information through the communication channel between the first terminal and the second terminal of the first user. The communication link can be a communication channel established between the first terminal and the second terminal, which can be based on the recommendation platform. For example, the recommendation platform establishes a communication channel between the first terminal and the second terminal in response to a call request from the target service personnel to the first user. Alternatively, after the recommendation platform sends the resource configuration information matched for the first user, recommendation information, and script information to the first terminal, the recommendation platform will automatically initiate a dual call to both the first and second terminals within a preset time to establish a communication channel between them.

[0027] Optionally, the resource information recommendation system further includes a service platform 104 (as in the service application described in the above embodiment). The service platform provides various resource configuration information for the target service area and the corresponding rights and interests for each resource configuration. Furthermore, enterprises providing services or goods can register as enterprise members on the service platform, and users requiring services or goods can register as user members on the service platform. The service platform can also establish a communication connection with the recommendation platform to synchronize various resource configuration information for the target service area, the configuration content of each resource configuration, enterprise information, and user information to the recommendation platform, so that the recommendation platform can store this information and perform actions based on this information. Figure 2aThe steps of the relevant embodiments. The users mentioned in the embodiments of this application primarily focus on B-end enterprises.

[0028] In this embodiment, the communication connection can be either a wireless communication connection or a wired communication connection. The specific communication connection method used depends on the implementation form of each component included in the resource information recommendation system. When the implementation form of the relevant component is physical (or physical), either a wireless communication connection or a wired communication connection can be used. When the implementation form of the relevant component is non-physical (virtual), a wireless communication connection can be used. Optionally, the wireless communication connection can be implemented through a mobile network. Correspondingly, the mobile network standard can be any one of 2G (GSM), 2.5G (GPRS), 3G (WCDMA, TD-SCDMA, CDMA2000, UTMS), 4G (LTE), 4G+ (LTE+), 5G, WiMax, or a new network standard that will emerge in the future. Furthermore, when the relevant components are located on the same local area network, wireless communication connections can also be implemented through Bluetooth, WiFi, infrared, Zigbee, or NFC.

[0029] It should be noted that, Figure 2a Related embodiments can be described around the system architecture of a resource information recommendation system, including a recommendation platform, a first terminal, a second terminal, and a service platform, but do not constitute a limitation on the technical solution of this application. Figure 2a The steps of the relevant embodiments can be found in the relevant descriptions of the embodiments below, and will not be repeated here.

[0030] Figure 2a This is a flowchart illustrating a resource information recommendation method provided for an exemplary embodiment of this application. Figure 2a The resource information recommendation methods include:

[0031] 201. In response to the resource information recommendation request submitted by the target service personnel in the target service domain, obtain the initial profile data and deep profile data of the first user. The initial profile data is obtained by user profile analysis based on the user information of the first user, and the deep profile data is obtained by user profile analysis based on the conversation content between the target service personnel and the first user.

[0032] 202. Based on the initial profile data and the deep profile data, conduct a potential resource demand analysis on the first user to obtain the potential multi-dimensional resource demand information of the first user;

[0033] 203. Input the initial profile data, deep profile data, multi-dimensional resource demand information, and various resource configuration information and their recommendation conditions into the AI ​​resource information recommendation model. Based on the recommendation conditions of the multi-dimensional resource demand information and resource configuration information, select the first target resource configuration information from the various resource configuration information and generate the recommendation reason information for the first target resource configuration information; analyze the language style information of the first user based on the initial profile data and deep profile data, and generate the speech information to recommend the first target resource configuration information to the first user based on the recommendation reason information; wherein, the AI ​​resource information recommendation model is obtained by fine-tuning the pre-trained model based on sample data of the target service domain;

[0034] 204. Send the first target resource configuration information, recommendation reason information, and script information to the first terminal of the target service personnel, so that the target service personnel can recommend the first target resource configuration information to the first user through the communication channel between the first terminal and the second terminal of the first user using the script information and recommendation reason information.

[0035] Figure 2b A flowchart illustrating another resource information recommendation method provided for an exemplary embodiment of this application. Figure 2c This is a flowchart illustrating yet another resource information recommendation method provided for an exemplary embodiment of this application. The following is in conjunction with... Figure 2b , Figure 2c The above technical solutions will be described in detail. Among them, Figure 2c This is a schematic diagram showing the modularization and hierarchical arrangement of various functions, specifically including a data layer, a model layer, and an application layer. The application layer includes a rule extraction module and a model fine-tuning module. The rule extraction module is used to select various rules, and the model fine-tuning module is used to fine-tune the model. The model layer includes the reasoning part of each model, and the application layer includes a sales assistance module that helps target service personnel recommend resource configuration information.

[0036] In this embodiment, the target service area is not specifically limited. The target service area could be, for example, recruitment services, housekeeping services, housing services, or food delivery services, etc.

[0037] In this embodiment, when a first user has a service requirement, fulfilling that requirement relies on resource configuration information provided by the service platform to provide the corresponding service to the first user. Taking a company as an example, the service could be a posting service. This posting service can fulfill the company's recruitment needs. The corresponding resource configuration information includes post resources. By publishing the company's recruitment needs through post resources, the company can complete the recruitment process. Post resources have resource types, including but not limited to: images, text, and videos. Based on this, service personnel in the relevant field can recommend resource configuration information suitable for the first user's service needs. For ease of description and distinction, the relevant field can be referred to as the target service field, and the service personnel as the target service personnel.

[0038] In this embodiment, target service personnel in the target service domain can register an account on the recommendation platform, becoming members of the recommendation platform. After logging into the service platform with their account, target service personnel can enjoy resource information recommendation services provided by the recommendation platform. Specifically, target service personnel can submit a resource information recommendation request for a first user on the recommendation platform, requesting the recommendation platform to select a first target resource configuration information suitable for the first user based on the resource information recommendation request, and to generate recommendation reason information and script information for the target service personnel to recommend the first target resource configuration information to the first user. Here, the first user is a B-end enterprise or C-end user of other applications associated with the recommendation platform. The resource information recommendation request includes the identification information of the first user, the identification information of the first terminal, and the identification information of the second terminal. The identification of the first user can be identity information with unique identification function, special code, code, etc. The identification information of the first terminal can be a terminal device identifier with unique identification function, special code, etc., but is not limited to these. The resource allocation information includes: the type and quantity of resource allocations, the geographical information that the resource allocation can serve, the user attributes that can be served (type, recruitment-related information), the resource name that reflects the resource allocation specifications (such as the number of people that can be recruited in the example below), the importance or priority of the resource, the price of the resource, etc. Taking posting as an example, the type of resource allocation can be the type of posting, and the quantity of resource allocations includes, but is not limited to: the number of posts that can be posted and the frequency.

[0039] This embodiment does not limit the specific implementation method of the target service personnel submitting resource information recommendation requests for the first user on the recommendation platform. For example, the recommendation platform can directly provide a configuration page for resource information recommendation requests. The configuration page includes at least information configuration items for the first user, information configuration items for the first terminal, and information configuration items for the second terminal. The identification information of the first user, the identification information of the first terminal, and the identification information of the second terminal can be configured in these information configuration items respectively. Alternatively, the recommendation platform and / or the service platform can pre-store the identification information of the first terminal associated with the identification information of the first user, and the recommendation platform and / or the service platform can pre-store the second identification information associated with the target service personnel. In response to the resource information recommendation request submitted by the target service personnel, the recommendation platform can obtain the identification information of the first terminal and the identification information of the second terminal from the local machine or the service platform in real time. For example, a recommendation platform provides a user information list (users of the service platform). This list includes information entries for each user, with each entry displaying the corresponding user's information. The target service personnel can filter out a first user based on the user information displayed in the list. This first user can be a user with potential demand for resource configuration information recommendations. Furthermore, when selecting the first user, information such as the user's historical order history, browsing history, and current browsing history can be used to determine the likelihood and degree of need for purchasing resource configuration information, thus identifying whether the user is a potential purchaser. The information entry corresponding to the first user may contain resources... The request control for information recommendation requests allows users to submit a resource information recommendation request first; or, in response to a filtering operation by a target service provider on a user information list, at least one selected user is added to the target service provider's candidate service list, which can hold information entries for multiple users; in response to a triggering operation by the target service provider on the information entry corresponding to the first user in the candidate service list, a resource information recommendation request is submitted for that first user. Each information entry is bound to the corresponding user's identification information. Submitting a resource information recommendation request for any user can automatically include the corresponding user's identification information, or it can automatically include the target service provider's identification information and the identification information of the first terminal device they are using. Therefore, prioritizing the recommendation of resource configuration information to potential first users can improve the conversion rate of resource configuration information.

[0040] In this embodiment of the application, in response to the resource information recommendation request, the recommendation platform can obtain the initial profile data and deep profile data of the first user. The initial profile data is obtained by user profile analysis based on the user information of the first user, and the deep profile data is obtained by user profile analysis based on the conversation content between the target service personnel and the first user.

[0041] In some embodiments, when registering as a member, users of the service platform fill in some basic information; or, users of the service platform may have the right to post a specified number of posts for free. Users can use this right to post posts, and the content of the posts usually includes the user's basic information, needs information, or resources that the user can provide. This information can be used as the user's user information to analyze the user's initial profile data.

[0042] Based on this, in response to resource information recommendation requests submitted by target service personnel in the target service domain, the system obtains the initial profile data and in-depth profile data of the first user, including: obtaining the user information of the first user based on the first user's identification information. The user information includes initial user data and initial service demand data. The initial user data contains the first user's basic information. When the first user is a C-end user, their basic information includes, but is not limited to: age, gender, occupation, educational background, marital status, income level, geographical location, historical information, and other registration information filled in by the first user when registering an account on the service platform. Historical information includes, but is not limited to: historical browsing history, historical posting history, and historical ordering history. When the user is a B-end enterprise, its basic information includes, but is not limited to: the enterprise's operating years, number of employees, business scope, income level, geographical location, historical resource information, and other registration information filled in by the first user when registering an account on the service platform; initial service demand data includes, but is not limited to: the demand information filled in by the first user when registering an account on the service platform, and the demand information contained in the post content published through the free posting rights provided by the service platform for the current demand. Taking the enterprise's recruitment demand as an example, the enterprise information (user information) can also be the customer type, and its demand information can be, for example, the recruitment scope, job positions, number of employees, recruitment budget, recruitment cycle, tool usage preferences, effect guarantee requirements, recruitment coins & tool discounts. Information such as card details, recruitment characteristics, and salary and benefits; furthermore, based on the identification information of the first terminal and the second terminal, a communication channel is established between the first terminal and the second terminal, and the conversation content generated during the conversation between the target service personnel and the first user based on the communication channel is obtained. The conversation content includes the first user's deep user data and deep service demand data. Deep user data refers to more detailed basic information related to the first user during the conversation, such as more detailed basic information related to the company's operating years, such as the company's founding date, or changes to the basic information listed above, or recently generated new basic information; in addition, more detailed basic information can also include... It includes basic information that was not previously obtained; deep service demand data refers to more detailed service demand information that was not included in the initial service demand data mentioned during the conversation. For example, more detailed service demand information related to the number of people to be recruited for a certain position could be the number of female employees and the number of male employees to be recruited. More detailed service demand information could also be change information of the demand information listed above, or new demand information that has recently been generated. In addition, more detailed service demand information can also include service demand information that was not previously obtained. Furthermore, the conversation content is input into a large language model to perform semantic understanding on the conversation content in order to obtain deep user data and deep service demand data related to the first user.User profiling is performed based on initial user data and initial service requirement data to obtain the initial user profile data; and user profiling is performed based on deep user data and deep service requirement data to obtain the deep user profile data. The initial and deep service requirement information describes the target services needed by the first user, and the execution of these target services depends on resource configuration information. After obtaining the initial user data and initial service requirement data, deep user data and deep service requirement data are obtained through sessions, which increases the richness of the relevant data for the first user. Based on this data, profiling analysis can improve the accuracy of the profile data. Deep profile data is more in-depth and specific than the initial profile data, and it includes some profile data not included in the initial profile data.

[0043] The large language model can be a multimodal large language model with deep semantic understanding and generation capabilities across modalities and languages. During user data extraction and profiling analysis, natural language processing can be used to analyze and classify user data, generating corresponding profiling data. Large language models can be, but are not limited to, Large Language Models (LLMs), GLM (Generalized Linear Model) series models, Qwen series models, etc. The large language model can be obtained by training an initial large language model using sample data from the profiling domain.

[0044] Optionally, the user information of the first user can be obtained based on the first user's identification information. This includes: if the recommendation platform locally stores the first user's user information synchronized with the service platform, then the user information of the first user can be directly obtained from the local storage based on the first user's identification information, thus facilitating the retrieval of this information. Alternatively, the user information of the first user can be obtained from the service platform in real time based on the first user's identification information. Since the first user's user information may be updated in real time on the service platform, in response to resource information recommendation requests submitted by target service personnel in the target service domain, obtaining the first user's user information from the service platform in real time based on the first user's identification information can obtain the latest user information and avoid missing updated user information.

[0045] Optionally, a communication channel is established between the first terminal and the second terminal based on the identification information of the first terminal and the second terminal, and the conversation content generated during the conversation between the target service personnel and the first user based on the communication channel is obtained. This includes: the recommendation platform responding to a call request from the target service personnel to the second user based on the outbound call control provided by the recommendation platform, making an outbound call to the second terminal to establish a communication channel between the first terminal and the second terminal. The recommendation platform can act as a relay station for the communication channel between the first terminal and the second terminal; in other words, the recommendation platform is the relay station for this communication channel, and the recommendation platform can obtain the conversation content of the conversation between the first terminal and the second terminal through this communication channel in real time. Alternatively, after the recommendation platform sends the resource configuration information, recommendation reason information, and script information matched for the first user to the first terminal, within a preset time, the recommendation platform will automatically initiate a dual call to the first terminal and the second terminal to establish a communication channel between the first terminal and the second terminal. The recommendation platform can act as a relay station for the communication channel between the first terminal and the second terminal; in other words, the recommendation platform is the relay station for this communication channel, and the recommendation platform can obtain the conversation content of the conversation between the first terminal and the second terminal through this communication channel in real time. The recommendation platform can obtain the conversation content between the target service personnel and the first user in real time through the above methods. This is beneficial for updating the first user's user information in a timely manner based on the conversation content, thereby improving the richness of user information.

[0046] Optionally, the conversation content is input into a large language model to perform semantic understanding, thereby obtaining deep user data and deep service demand data related to the first user. This includes: inputting the conversation content into the large language model, performing semantic understanding to obtain semantic information of the conversation content; classifying the semantic information of the conversation content from user data dimensions and demand data dimensions to obtain semantic information of user data dimensions and demand data dimensions respectively; and performing semantic transformation on the semantic information of user data dimensions and demand data dimensions respectively to obtain deep user data and deep service demand data related to the first user. Extracting deep user data and deep service demand data separately through semantic understanding can improve the efficiency and accuracy of data extraction, thereby improving the efficiency of resource information generation.

[0047] It should be noted that during the process of acquiring the session content, the large language model can be used directly to perform semantic understanding on the session content, thereby obtaining in-depth user data and in-depth service demand data related to the first user. Compared with performing semantic understanding on the session content after it has been acquired, this improves the efficiency of data analysis, which in turn improves the efficiency of resource information generation.

[0048] It should also be noted that communication through established channels can occur multiple times, and the session content can be the sum of the content from these multiple conversations. During each return to the session content, semantic understanding can be performed to obtain in-depth user data and in-depth service requirement data relevant to the first user in that session. Furthermore, resource information can be generated based on each data point, and corresponding resource configuration information can be recommended to the user until the user is satisfied. Alternatively, after the obtained in-depth user data and in-depth service requirement data from multiple sessions related to the first user meet the requirements of the target service personnel, each data point can be integrated, and resource information can be generated based on the integrated data.

[0049] Optionally, user profiling analysis is performed based on initial user data and initial service demand data to obtain initial user profile data. This includes: inputting the initial user data and initial service demand data into a target profile analysis model; performing semantic understanding on the initial user data and initial service demand data from a profile dimension to obtain a target profile semantic information set; performing semantic classification on the profile semantic information in the target profile semantic information set to obtain multiple target profile semantic categories; and mapping profile information across the multiple target profile semantic categories based on a pre-learned mapping relationship between profile semantic categories and profile information in the target domain to obtain the initial user profile data. Classifying profile semantic categories using the semantic information of the initial user data and obtaining the initial user profile data based on the mapping relationship between profile semantic categories and profile information can improve the accuracy and efficiency of the initial profile data analysis.

[0050] Optionally, user profiling analysis based on deep user data and deep service demand data can be performed to obtain deep user profile data for the first user. This includes: inputting deep user data and deep service demand data into a target profile analysis model; performing semantic understanding on the deep user data and deep service demand data from a profile dimension to obtain a target profile semantic information set; performing semantic classification on the profile semantic information in the target profile semantic information set to obtain multiple target profile semantic categories; and mapping profile information across the multiple target profile semantic categories based on a pre-learned mapping relationship between profile semantic categories and profile information to obtain the deep user profile data for the first user. Classifying profile semantic categories using the semantic information of user deep data and obtaining the deep user profile data for the first user based on the mapping relationship between profile semantic categories and profile information can improve the accuracy and efficiency of deep profile data analysis.

[0051] In this embodiment, after obtaining the initial and deep profile data of the first user, potential resource needs analysis can be performed based on these data to obtain multi-dimensional resource needs information. This multi-dimensional resource needs information can be understood as multi-dimensional business opportunity information. From the resource provider's perspective, this information reflects the user's business opportunities with the resource provider from multiple dimensions. Combining the initial and deep profile data to analyze the first user's potential resource needs improves the accuracy of this information. Furthermore, detailed analysis of the first user's potential resource needs from multiple dimensions allows for precise acquisition of resource needs information for each dimension. This facilitates matching the first user with more suitable resource configurations based on accurate multi-dimensional resource needs information, thereby improving the conversion rate of resource configuration information. The multi-dimensional resource needs information is detailed information about the first user's potential business needs and cooperation possibilities, mined from different angles and levels. This includes, but is not limited to: budget level, duration of resource configuration information needs, frequency of resource configuration information usage, service function requirements, brand loyalty, and willingness to cooperate. By conducting detailed analysis of these multi-dimensional resource demand information, we can provide users with more accurate and suitable resource allocation information, thereby improving the conversion rate of resource allocation information and user satisfaction.

[0052] In some embodiments, a potential resource demand analysis is performed on a first user based on initial profile data and deep profile data to obtain the first user's potential multi-dimensional resource demand information. This includes: inputting the initial profile data and deep profile data into a target resource demand analysis model; performing semantic understanding on the initial profile data and deep profile data from the resource demand dimension to obtain a target resource demand semantic information set, which includes the resource demand semantic information contained in the initial profile data and deep profile data, as well as the resource demand semantic information contained in both; performing semantic classification on the resource demand semantic information in the target resource demand semantic information set to obtain multiple target semantic categories; and mapping the resource demand information of the multiple target semantic categories based on the pre-learned mapping relationship between semantic categories and resource demand information in the target service domain to obtain the first user's potential multi-dimensional resource demand information. Combining the initial profile data and deep profile data to analyze the first user's potential resource demand can improve the accuracy of multi-dimensional resource demand information, thereby helping to improve the suitability of recommended resource configuration information with the first user.

[0053] Further optional, such as Figure 2eAs shown, the target resource demand analysis model described above can be trained in the following way: A sample dataset from the target service domain is obtained, including multiple sample profile data, sample resource demand semantic information corresponding to each sample profile data, and sample semantic category corresponding to each sample resource demand semantic information; multiple sample profile data are input into the initial resource demand analysis model, and semantic understanding of the multiple sample profile data is performed from the resource demand dimension to obtain intermediate-state resource demand semantic information; the intermediate-state resource demand semantic information is semantically classified to obtain multiple intermediate-state semantic categories, and the mapping relationship between the intermediate-state semantic categories and the intermediate-state resource demand semantic information is learned; based on the mapping relationship, resource demand information is mapped to the multiple intermediate-state semantic categories to obtain intermediate-state resource demand information; the model loss function is calculated based on the intermediate-state resource demand semantic information, sample resource demand semantic information, and intermediate-state semantic categories; if the model loss function does not meet the model training termination condition, the initial resource demand analysis model is continued to be trained until the model loss function meets the model training termination condition, thus obtaining the target resource demand analysis model. After obtaining the target resource demand analysis model, its performance is evaluated. If the evaluation is satisfactory, the model is deployed to the appropriate operating environment. Furthermore, the model can be optimized in real-time based on usage data to improve its performance. Specifically, the initial resource demand analysis model is selected from the models to be pre-trained based on the target service domain and trained using training samples from the pre-training data preparation library. Fine-tuning of the model using sample data from the target service domain improves the accuracy of the target resource demand model in analyzing resource demands within the target service domain, thereby enhancing the accuracy of the first user's potential multi-dimensional resource demand information.

[0054] Optionally, the model loss function is calculated based on the intermediate state resource demand semantic information and the sample resource demand semantic information, as well as the intermediate state semantic category and the sample semantic category. This includes: calculating a first loss function between the intermediate state resource demand semantic information and the sample resource demand semantic information, and calculating a second loss function between the intermediate state semantic category and the sample semantic category, both of which are used as the model loss function; or, calculating the first loss function between the intermediate state resource demand semantic information and the sample resource demand semantic information, and calculating the second loss function between the intermediate state semantic category and the sample semantic category, and then performing a weighted sum of the first and second loss functions to obtain the model loss function. By using multiple loss functions individually or in combination, the model can be fine-tuned, thereby improving the model's accuracy.

[0055] In this embodiment, each type of resource configuration information has its own applicable conditions, which include, but are not limited to, the user attributes and target service attributes to which the resource configuration information applies. The target service attributes include multiple attributes of the target service and different attribute values ​​for each attribute. User attributes describe user characteristics, such as user ID, user type, and user level. The target service attributes describe service characteristics, such as service type, service level, and service availability. Attribute values ​​refer to the specific values ​​corresponding to each attribute, used to specifically describe a certain characteristic of the user or service. More specifically, the applicable conditions for each resource configuration information refer to a series of regulations that any user must meet when using the resource configuration information. These applicable conditions include which users can enjoy the various rights and / or benefits provided by the resource configuration information. Furthermore, the applicable conditions for the resource configuration information also clarify the validity period, scope of use, and usage restrictions of the resource configuration information to ensure that users can fully understand and comply with the relevant regulations during use, thereby protecting the rights and interests of both parties.

[0056] Based on the applicable conditions of each type of resource configuration information described above, to improve the suitability of the resource configuration information selected for the first user, multi-dimensional resource demand information and various resource configuration information and their recommendation conditions can be input into the AI ​​resource information recommendation model. Based on the multi-dimensional resource demand information and the recommendation conditions of the resource configuration information, the model selects the first target resource configuration information from the various resource configuration information and generates a recommendation reason for the first target resource configuration information. Specifically, the recommendation conditions for resource configuration information refer to the methods used to accurately recommend the most suitable resource configuration information to the user based on the first user's multi-dimensional resource demand information. These resource configuration information rules aim to maximize user satisfaction while improving the conversion rate and user retention rate of resource configuration information.

[0057] In some embodiments, the initial profile data, the deep profile data, multi-dimensional resource requirement information, and various resource configuration information and their recommendation conditions are input into an AI resource information recommendation model. Based on the recommendation conditions of the multi-dimensional resource requirement information and resource configuration information, target resource configuration information is selected from various resource configuration information. This includes: inputting the initial profile data, the deep profile data, the multi-dimensional resource requirement information, various resource configuration information and their recommendation conditions, and the applicable conditions of various resource configuration information into the AI ​​resource information recommendation model; performing semantic matching between the multi-dimensional resource requirement information and the applicable conditions of each resource configuration information to determine the target applicable conditions that semantically match the multi-dimensional resource requirement information; selecting resource configuration information with the target applicable conditions from the various resource configuration information as candidate resource configuration information; and selecting the first target resource configuration information from the candidate resource configuration information based on the resource configuration information recommendation conditions and the multi-dimensional resource requirement information. Semantically matching the resource requirement information of each dimension with the applicable conditions of each resource configuration information improves the accuracy of semantic matching, thereby improving the fit between the first target resource configuration information and the first user.

[0058] It should be noted that the applicable conditions for each resource allocation information and the recommended conditions for resource allocation information are different rules, and their content and functions are different.

[0059] To provide users with a wide range of choices and meet the needs of different users, and to further improve the adaptability between the first target service and the first user, in this embodiment, there can be multiple recommendation conditions for resource configuration information. Based on the recommendation conditions of resource configuration information and multi-dimensional resource demand information, the first target resource configuration information is selected from the candidate resource configuration information. This includes: grouping multiple resource configuration information based on the recommendation conditions of each resource configuration information to obtain an initial resource configuration information set that is adapted to the recommendation conditions of each resource configuration information; for each initial resource configuration information set, analyzing the local adaptability of each resource configuration information in the initial resource configuration information set with the resource demand information of each dimension in the multi-dimensional resource demand information from the perspective of service content; calculating the overall adaptability based on the local adaptability of each resource configuration information with the resource demand information of each dimension in the multi-dimensional resource demand information; and taking the resource configuration information with an overall adaptability greater than the corresponding threshold as the first target resource configuration information under the recommendation conditions of the resource configuration information.

[0060] Below is an example of resource configuration information:

[0061] "Resource Allocation Information Number - Nationwide Recruitment - Version Model - Standard Edition Quarterly Member Resource Allocation Information (Resource Allocation Information Name)"

[0062] Customer Analysis:

[0063] 1. Customer type: Directly recruited customers

[0064] 2. Recruitment Scope: Nationwide

[0065] 3. Job openings: (1,4);

[0066] 4. Number of positions to be filled: (2, 6);

[0067] 5. Recruitment budget (1000-2000);

[0068] 6. Recruitment cycle: Short-term recruitment, seasonal recruitment;

[0069] 7. Item usage preferences: Exposure, efficiency;

[0070] 8. Results guarantee requirements: None;

[0071] 9. Recruitment Coins & Item Discount Cards: None

[0072] 10. Recruitment characteristics: clients who are eager to try something new, have very low budgets, and are concerned about the platform's effectiveness.

[0073] Selling points of resource allocation information: Quarterly resource allocation information can be posted 3 times, with super pinned posts to increase exposure. Combined with talent radar and high-speed phone cards, it enables efficient screening and contacting of job seekers.

[0074] Precautions:

[0075] 1. Directly recruited clients who are unable to sign up for monthly personal value-added resource allocation information (local) may consider purchasing quarterly membership resource allocation information (nationwide).

[0076] 2. Popular job postings are limited to 5 per category: food delivery rider, delivery driver, ride-hailing driver, business driver, chauffeur driver, taxi driver, shuttle bus driver, and passenger transport driver.

[0077] 3. The "Super Top" section in the Quarterly Member Resource Allocation Information (Nationwide) cannot be used for the aforementioned 8 categories.

[0078] In some embodiments, the multi-dimensional resource requirement information includes: attribute information of the first user, budget requirement level information, and service requirement information. The service requirement information includes: basic service requirements, core service requirements, and service requirement duration. The resource allocation information recommendation criteria include optimal resource allocation information recommendation criteria, medium resource allocation information recommendation criteria, and minimum resource allocation information recommendation criteria. The optimal resource allocation information recommendation criteria are used to determine resource allocation information that meets the high budget level and service requirement information. The medium resource allocation information recommendation criteria are used to determine resource allocation information that meets the low budget level, meets the core service requirements, and meets the service requirement duration. The minimum resource allocation information recommendation criteria are used to determine resource allocation information that meets the basic service requirements and meets the service requirement duration. For example, the different first target resource allocation information templates recommended for the first user are roughly as follows:

[0079] Optimal resource allocation information: high value, high matching degree (sufficient budget / matching needs);

[0080] Medium-level resource allocation information: high cost-effectiveness, covering core needs (limited budget);

[0081] Guaranteed resource allocation information: covers basic services and has low barriers to entry (new clients / short-term cooperation).

[0082] Based on the aforementioned multi-dimensional resource requirement information, to further improve the fit between the primary target service and the primary user, for each initial resource configuration information set, the local fit between each resource configuration information in the initial resource configuration information set and the resource requirement information in each dimension of the multi-dimensional resource requirement information is analyzed from the service content dimension. The overall fit is calculated based on the local fit of each resource configuration information and the resource requirement information in each dimension of the multi-dimensional resource requirement information. Resource configuration information with an overall fit greater than a corresponding threshold is used as the primary target resource configuration information under the recommended conditions. This includes: for the initial resource configuration information set under the recommended conditions of the optimal resource configuration information, analyzing multiple first local fits of each resource configuration information in the initial resource configuration information set in the dimensions of budget requirement level, basic service requirement, core service requirement, and service requirement duration; calculating the first overall fit based on the multiple first local fits; and selecting resource configuration information with a first overall fit greater than a first threshold. As the first target resource allocation information under the optimal resource allocation information recommendation condition; for the initial resource allocation information set under the medium resource allocation information recommendation condition, analyze the multiple second local fit degrees of each resource allocation information in the initial resource allocation information set in the dimensions of budget demand level, core service demand, and service demand duration; based on the multiple second local fit degrees, calculate the second overall fit degree; the resource allocation information corresponding to the second overall fit degree greater than the second threshold is used as the first target resource allocation information under the medium resource allocation information recommendation condition; for the initial resource allocation information set under the minimum resource allocation information recommendation condition, analyze the multiple third local fit degrees of each resource allocation information in the initial resource allocation information set in the dimensions of budget demand level, basic service demand, and service demand duration; based on the multiple third local fit degrees, calculate the third overall fit degree; the resource allocation information corresponding to the third overall fit degree greater than the third threshold is used as the first target resource allocation information under the minimum resource allocation information recommendation condition.

[0083] Taking a company as the primary user and its target service area as recruitment services as an example, the company's information is as follows:

[0084] * Company Name: A car rental company in a certain location;

[0085] * Date of Establishment: **Year** **Month**;

[0086] * Staff size: 10-50 people;

[0087] * Main business: New energy vehicle leasing;

[0088] * Job openings: Leasing consultant, vehicle dispatcher, etc.

[0089] Recruitment pain points:

[0090] * Low resume matching rate: When recruiting through traditional classifieds platforms, the resume matching rate is less than 20%.

[0091] * Core issues: Insufficient industry awareness and lack of local resources.

[0092] Based on the above enterprise information, the optimal resource allocation, medium resource allocation, and minimum resource allocation information are determined as follows:

[0093] Optimal resource allocation information:

[0094] Is it a primary promotion: Yes. City Tier: Tier 1 cities. Business Type: New Sign-up / Renewal. Resource Configuration Information Number: PKG_8180. Recommended Resource Configuration Information Name: Resource Configuration Information Number - Full Network Recruitment 3.0 Standard Version Annual Membership Resource Configuration Information.

[0095] Customer Analysis:

[0096] 1. Customer type: Directly recruited customers (first-tier clients);

[0097] 2. Recruitment Scope: Nationwide;

[0098] 3. Job openings: (3, 8);

[0099] 4. Number of positions to be filled: (10, 25)

[0100] 5. Recruitment budget (8000 or more);

[0101] 6. Recruitment cycle: Recruiting continuously throughout the year;

[0102] 7. Item usage preference: Flexible;

[0103] 8. Results guarantee requirements: None;

[0104] 9. Recruitment Coins & Item Discount Cards: 5500 Recruitment Coins, 8 discount cards at 65% off;

[0105] 10. Recruitment Characteristics: Moderate budget, long-term recruitment needs. Resource Allocation Information Highlights: Includes 8 job postings, 5500 recruitment coins (equivalent to cash on the platform, more cost-effective than 6180), 8 discount cards for a 65% discount on featured positions, etc. Notes: Limited to 5 posts across 8 popular job categories: Food Delivery Rider, Delivery / Delivery Personnel, Ride-Hailing Driver, Business Driver, Designated Driver, Taxi Driver, Shuttle Bus Driver, Passenger Transport Driver.

[0106] Medium resource allocation information:

[0107] Is it a primary promotion: Yes. City tier: Nationwide. Business type: New signing / Online promotion. Resource configuration information number: PKG_4980. Recommended resource configuration information name: Resource configuration information number - Nationwide Recruitment 3.0 Standard Version Annual Membership Resource Configuration Information.

[0108] Customer Analysis:

[0109] 1. Customer type: Directly recruited customers;

[0110] 2. Recruitment Scope: Nationwide;

[0111] 3. Job openings: (3, 8);

[0112] 4. Number of positions to be filled: (6, 20)

[0113] 5. Recruitment budget (3000-5000);

[0114] 6. Recruitment cycle: Recruiting continuously throughout the year;

[0115] 7. Item usage preference: Flexible;

[0116] 8. Results guarantee requirements: None;

[0117] 9. Recruitment Coins & Item Discount Cards: 3600 Recruitment Coins, 8 discount cards at 65% off;

[0118] 10. Recruitment Characteristics: Limited budget, long-term recruitment needs. Resource Allocation Highlights: Includes 8 job postings, 3600 recruitment coins (equivalent to cash on the platform), 8 discount cards for 65% off featured positions, etc., achieving higher exposure. Notes: Postings are limited to 5 out of 8 popular job categories: Food Delivery Rider, Delivery / Delivery Personnel, Ride-Hailing Driver, Business Driver, Designated Driver, Taxi Driver, Shuttle Bus Driver, Passenger Transport Driver.

[0119] Guaranteed resource allocation information:

[0120] Is it a primary promotion: Yes. City tier: Nationwide. Business type: New signing / Online promotion. Resource configuration information number: PKG_3980. Recommended resource configuration information name: Resource configuration information number - Nationwide Recruitment 3.0 Standard Version Annual Membership Resource Configuration Information.

[0121] Customer Analysis:

[0122] 1. Customer type: Directly recruited customers;

[0123] 2. Recruitment Scope: Nationwide;

[0124] 3. Job openings: (3, 8);

[0125] 4. Number of positions to be filled: (5, 15)

[0126] 5. Recruitment budget (3000-5000);

[0127] 6. Recruitment cycle: Recruiting continuously throughout the year;

[0128] 7. Item usage preference: Flexible;

[0129] 8. Results guarantee requirements: None;

[0130] 9. Recruitment Coins & Item Discount Cards: 2400 Recruitment Coins, 8 discount cards with a 68% discount rate;

[0131] 10. Recruitment Characteristics: Low budget, long-term recruitment needs. Resource Allocation Information Selling Points: Lowest-priced nationwide resource allocation information, including 8 job listings, 2400 recruitment coins, and 8 68% off item discount cards; more flexible combination usage. Notes: Limited to 5 posts in 8 popular job categories: Food Delivery Rider / Rider, Delivery / Delivery Personnel, Ride-Hailing Driver, Business Driver, Designated Driver, Taxi Driver, Shuttle Bus Driver, Passenger Transport Driver.

[0132] Furthermore, the AI ​​resource information recommendation model can also generate recommendation reason information for the first target resource configuration information, so that the target service personnel can recommend the first target resource configuration information to the first user based on the recommendation reason information. The recommendation reason information refers to the explanatory text used by the target service personnel to recommend the first target resource configuration information to the first user.

[0133] In one optional embodiment, generating recommendation reason information for the first target resource configuration information includes: generating service requirement-type recommendation reason information based on the service requirements pointed to by the multi-dimensional resource requirement information and the configuration services represented by the configuration content of the first target resource configuration information; generating budget-type recommendation reason information based on the budget requirements pointed to by the multi-dimensional resource requirement information and the configuration prices represented by the configuration content of the first target resource configuration information; and integrating the service requirement-type and budget-type recommendation reason information to generate recommendation reason information for recommending the first target resource configuration information to the first user. The recommendation reason information ensures that the resource configuration matches the user's needs, enhances persuasiveness, improves the conversion rate of the resource configuration information, and enhances the user experience.

[0134] Furthermore, to improve the conversion rate of resource configuration information recommended by target service personnel to the first user, the AI ​​resource information recommendation model can also analyze the language style information of the first user based on the initial profile data and the deep profile data. Based on the language style information, it generates script information to recommend the first target resource configuration information to the first user using the recommendation reason information. This allows target service personnel to use the script techniques provided by the script information to recommend the first target resource configuration information to the first user using the recommendation reason information. Language style information refers to a series of features and attributes manifested in textual, spoken, or other forms of language communication. These features and attributes reflect the speaker's or author's personality, emotions, intentions, context, and cultural background. This information helps target service personnel better understand the first user's language habits, thereby increasing the first user's acceptance of the recommendation process. Script information refers to the script techniques used to recommend the first target resource configuration information to the first user. A script template library corresponds to the script information, and suitable script templates can be selected from the template library according to specific needs (such as user type) to generate script information based on the script model. The script templates are used to standardize the content and structure of the script information.

[0135] In one optional embodiment, analyzing the language style information of the first user based on the initial profile data and the deep profile data includes: inputting the initial profile data and the deep profile data into a pre-trained language style analysis model; extracting language style-related features from the initial profile data and the deep profile data, including but not limited to word frequency, sentence structure, sentiment tendency, semantic density, etc.; and analyzing the language style information of the first user based on the extracted language style-related features, including but not limited to formal and informal style, spoken and written style, intensity of emotional expression, etc.

[0136] In one optional embodiment, generating script information to recommend first target resource configuration information to the first user based on the first user's language style information and recommendation reason information includes: generating script information based on the language style information and recommendation reason information for recommending the first target resource configuration information to the first user. This allows for more accurate fulfillment of user needs, improving user satisfaction and the conversion rate of resource configuration information. More specifically, script information can be generated based on the first user's language style information and the recommendation format corresponding to the recommendation reason information, ensuring that the script information not only matches the first user's language style but also the recommendation format of the recommendation reason information, thereby improving the first user's acceptance of the first target resource configuration information. Here, the recommendation format refers to the presentation method and structure of the recommended content included in the recommendation reason information, which determines how the recommendation reason information is organized and displayed to the user. The recommendation format can be adjusted according to different user needs, scenarios, and language styles to improve user acceptance and satisfaction.

[0137] Furthermore, to enhance the first user's trust in the resource allocation information recommendation service, case studies of users who have purchased resource allocation information and provided them with positive feedback can be selected from historical users. Specifically, a second user can be identified from the historical users. This second user's multi-dimensional resource requirement information must match the first user's multi-dimensional resource requirement information to a level greater than a fourth threshold, and their evaluation of the second target resource allocation information recommended and purchased by the target service provider must be higher than a fifth threshold. This step can be considered as calculating the similarity between the multi-dimensional resource requirement information of the first user and the second user.

[0138] In some optional embodiments, determining a second user from historical users includes: obtaining historical evaluation information, which includes multi-dimensional resource requirement information of multiple historical users, purchased second target resource configuration information and configuration content, evaluation information and corresponding evaluation level of each historical user on the purchased second target resource configuration information; based on the multi-dimensional resource requirement information of multiple historical users, determining a target historical user whose multi-dimensional resource requirement information matches the first user's multi-dimensional resource requirement information more than a fourth threshold; determining a second user from the target historical user whose evaluation level of the purchased second target resource configuration information is higher than a fifth threshold; and pushing the historical evaluation information of the second user to a second terminal so that the first user can understand the historical evaluation information of the second target resource configuration information recommended by the target service personnel to the second user.

[0139] Accordingly, generating recommendation reason information for the first target resource configuration information includes: generating service demand-type recommendation reason information based on the service demands pointed to by the multi-dimensional resource demand information and the resource configuration information service represented by the configuration content of the first target resource configuration information; generating budget-type recommendation reason information based on the budget demands pointed to by the multi-dimensional resource demand information and the resource configuration information price represented by the configuration content of the first target resource configuration information; generating historical evaluation-type recommendation reason information based on the second user's multi-dimensional demand information, the purchased second target resource configuration information, and the historical evaluation information pointed to by the evaluation level of the purchased second target service; integrating the service demand-type recommendation reason information, the budget-type recommendation reason information, and the historical evaluation-type recommendation reason information to generate recommendation reason information for recommending the first target resource configuration information to the first user; and analyzing the language style information of the first user based on the initial profile data and the deep profile data, and generating persuasive language information to recommend the first target resource configuration information to the first user based on the language style information.

[0140] Furthermore, such as Figure 2e As shown, after obtaining the first target resource configuration information, recommendation reason information, and script information, resource information can be generated based on the preset resource information structure, and the structured resource information can be sent to the target service personnel.

[0141] Taking the aforementioned car rental company recruitment as an example, the reasons for service needs are as follows: Your company is currently in a phase of rapid development, recruiting for positions such as rental consultants and vehicle dispatchers, and faces core issues such as low resume matching, insufficient industry knowledge, and a lack of local resources. Addressing these pain points, the resource allocation information we recommend provides nationwide recruitment services, helping you expand your recruitment scope and attract more professional talent. Among them, the Standard Annual Membership Resource Allocation Information (Optimal Resource Allocation Information) supports continuous recruitment throughout the year, meeting your medium- to long-term recruitment needs, and provides 8 job listings to cover your current recruitment requirements. The reasons for budget consideration are as follows: Based on your recruitment budget, we offer resource allocation information options at different price points: Standard Annual Membership Resource Allocation Information (Optimal Resource Allocation Information): Suitable for companies with a moderate budget and long-term recruitment needs, this resource allocation information costs 8180 yuan, providing 5500 recruitment coins, equivalent to cash, offering high cost-effectiveness, and includes 8 65% discount cards that can be used to purchase value-added services such as featured listings to further enhance recruitment effectiveness. Standard Annual Membership Resource Allocation Information (Medium Resource Allocation Information): Suitable for companies with a budget between 3000-5000 RMB. The resource allocation information costs 4980 RMB, providing 3600 recruitment coins (equivalent to cash), and includes 8 65% discount cards. It can meet your recruitment needs while achieving higher exposure. Standard Annual Membership Resource Allocation Information (Minimum Guarantee Resource Allocation Information): Suitable for companies with a lower budget. The resource allocation information costs 3980 RMB, providing 2400 recruitment coins, and includes 8 68% discount cards. This is the lowest-priced nationwide resource allocation information, offering more flexibility when used in combination, helping you achieve your recruitment goals within a limited budget. Historical evaluations are based on the following: We referenced the multi-dimensional needs, purchased resource allocation information, and ratings of resource allocation information from companies of similar size and industry (such as second-hand users). These companies generally expressed satisfaction with the recruitment results of the resource allocation information, believing it effectively improved resume matching and recruitment efficiency. For example, companies that purchased medium-level resource allocation information reported that it saved them time and costs during the recruitment process, while attracting more high-quality candidates. Therefore, we believe that this resource allocation information can also meet your recruitment needs and help you solve your current recruitment pain points.

[0142] Accordingly, the script information is as follows: This company is a car rental company and values ​​drivers' service capabilities and friendliness. When communicating with users, pay attention to the following aspects: Friendliness: Use a friendly and approachable language style to shorten the distance with customers; Cultural Adaptability: Use a language style that conforms to business communication habits, while taking into account the company's industry characteristics and cultural background; Guidance: Guide customers to focus on recommended resource allocation information by gradually introducing its features and advantages; Targeted Approach: Closely combine the customer's background, pain points, and needs to provide personalized resource allocation information recommendations; Flexibility: Offer multiple resource allocation information options to meet different budget requirements; Professionalism: Use professional recruitment service terminology to demonstrate professionalism.

[0143] Based on the above recommendation reasons and script information, the target service personnel recommend the following text content to the user regarding resource configuration information:

[0144] "Dear Car Rental Company Manager:"

[0145] Hello! I am a service representative from the recruitment service platform [Platform Name], and it is a great honor to contact your company. We understand that your company is rapidly developing in the new energy vehicle rental sector and is currently facing some recruitment challenges, particularly in resume matching and local resources. We deeply understand that recruitment is crucial for a company's development, and therefore we hope to provide you with some practical solutions.

[0146] Before contacting you, we conducted an in-depth analysis of your company's business scope, recruitment needs, and budget. Based on this information, we have carefully selected three recruitment resource configurations to meet your different needs.

[0147] First, we recommend the optimal resource allocation information. This information is ideal for companies like yours with medium- to long-term recruitment needs. It supports continuous recruitment throughout the year, offering 8 job postings to meet your current recruitment requirements. The resource allocation information is priced at 8180 yuan, including 5500 recruitment coins, which can be used as cash, offering excellent value. In addition, we are also providing you with 8 65% discount cards, which you can use to purchase value-added services such as featured placement to further enhance recruitment effectiveness.

[0148] If you have a tighter budget, we also offer two resource allocation options. The medium-level resource allocation costs 4980 yuan, provides 3600 recruitment coins (equivalent to cash), and includes eight 65% discount cards, meeting your recruitment needs while achieving higher exposure. The guaranteed minimum resource allocation is the lowest-priced nationwide option, costing 3980 yuan, providing 2400 recruitment coins, and including eight 68% discount cards for promotional items. This option offers greater flexibility and can help you achieve your recruitment goals within a limited budget.

[0149] We also referenced feedback from companies of similar size and industry regarding these resource allocation information. Companies that purchased medium-level resource allocation information reported that it saved them time and costs during the recruitment process, while attracting more high-quality candidates. Therefore, we believe this resource allocation information can also meet your recruitment needs and help you address your current recruitment pain points.

[0150] During our communication with you, we noticed your company's emphasis on recruitment effectiveness and reasonable budget planning. We believe that through our professional services and carefully designed resource allocation information, we can bring you satisfactory recruitment results. If you have any questions about resource allocation information, or need further details on other resource allocation methods, please feel free to contact us. We look forward to cooperating with your company and contributing to your business development!

[0151] In the technical solutions provided by the above embodiments of this application, an AI resource information recommendation model is used to determine the first target resource configuration information from a variety of resource configuration information based on the recommendation conditions of the first user's potential multi-dimensional resource demand information and resource configuration information, so as to quickly and accurately match the user with suitable resource configuration information. In addition, the language style information of the first user can be analyzed based on the initial and deep profile data of the first user, and recommendation reason information and wording information can be generated to recommend the first target resource configuration information to the first user. The target service personnel can use the wording skills provided by the wording information to recommend the first target resource configuration information to the first user with the recommendation reason information, which improves the persuasiveness of the recommended resource configuration information, and also improves the user's acceptance of the recommendation process by adapting the language style to the first user, thereby improving the conversion rate of resource configuration information.

[0152] Figure 3 A schematic diagram of the structure of an electronic device provided as an exemplary embodiment of this application. For example... Figure 3 As shown, it includes: a memory 30a and a processor 30b; the memory 30a is used to store computer programs; the processor 30b is coupled to the memory 30a and is used to execute the computer programs to perform the steps in the above method embodiments.

[0153] Furthermore, such as Figure 3 As shown, the server also includes other components such as a communication component 30c, a display 30d, a power supply component 30e, and an audio component 30f. Figure 3 The diagram only shows some components and does not mean that the electronic device includes only these components. Figure 3 The components shown.

[0154] The detailed implementation methods and beneficial effects of the electronic devices provided in this application have been described in detail in the foregoing embodiments, and will not be elaborated further here.

[0155] Exemplary embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps in the above-described method embodiments.

[0156] An exemplary embodiment of this application also provides a computer program product comprising a computer program / instructions that, when executed by a processor, enable the processor to perform the steps described in the above method embodiments.

[0157] The aforementioned memory can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random-Access Memory (SRAM), Electrically Erasable Programmable Read Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0158] The aforementioned communication components are configured to facilitate wired or wireless communication between the device containing the communication components and other devices. The device containing the communication components can access wireless networks based on communication standards, such as WiFi, 2G, 3G, 4G / LTE, 5G, or combinations thereof. In one exemplary embodiment, the communication components receive broadcast signals or broadcast-related information from an external broadcast management system via a broadcast channel. In one exemplary embodiment, the communication components also include a Near Field Communication (NFC) module to facilitate short-range communication. For example, the NFC module may be implemented based on Radio Frequency Identification (RFID), Infrared Data Association (IrDA), Ultra Wide Band (UWB), Bluetooth (BT), and other technologies.

[0159] The aforementioned display includes a screen, which may include a Liquid Crystal Display (LCD) and a Touch Panel (TP). If the screen includes a Touch Panel, the screen can be implemented as a touchscreen to receive input signals from the user. The Touch Panel includes one or more touch sensors to sense touches, swipes, and gestures on the Touch Panel. The touch sensors can sense not only the boundaries of touch or swipe actions but also the duration and pressure associated with the touch or swipe operation.

[0160] The aforementioned power supply components provide power to various components within the device in which they reside. These power supply components may include a power management system, one or more power sources, and other components associated with generating, managing, and distributing power to the device in which they reside.

[0161] The aforementioned audio component can be configured to output and / or input audio signals. For example, the audio component includes a microphone (MIC) configured to receive external audio signals when the device containing the audio component is in an operating mode, such as call mode, recording mode, or voice recognition mode. The received audio signals can be further stored in memory or transmitted via a communication component. In some embodiments, the audio component also includes a speaker for outputting audio signals.

[0162] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-readable storage media (including, but not limited to, disk storage, compact disc read-only memory (CD-ROM), optical storage, etc.) containing computer-usable program code.

[0163] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations 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, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0164] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0165] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0166] In a typical configuration, a computing device includes one or more processors (Central Processing Unit, CPU), input / output interfaces, network interfaces, and memory.

[0167] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0168] Computer-readable media include both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. 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 random access 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 technologies, CD-ROM, Digital Video Disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0169] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0170] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A resource information recommendation method, characterized in that, include: In response to a resource information recommendation request submitted by a target service personnel in the target service domain, initial profile data and deep profile data of a first user are obtained. The initial profile data is obtained by user profile analysis based on the user information of the first user, and the deep profile data is obtained by user profile analysis based on the conversation content between the target service personnel and the first user. Based on the initial profile data and the deep profile data, a potential resource demand analysis is performed on the first user to obtain the first user's potential multi-dimensional resource demand information. The initial profile data, the deep profile data, the multi-dimensional resource demand information, and the various resource configuration information and their recommendation conditions are input into the AI ​​resource information recommendation model. Based on the recommendation conditions of the multi-dimensional resource demand information and the resource configuration information, the first target resource configuration information is selected from the various resource configuration information and the recommendation reason information of the first target resource configuration information is generated. Based on the initial profile data and the deep profile data, the language style information of the first user is analyzed, and based on the language style information, the wording information for recommending the first target resource configuration information to the first user with the recommendation reason information is generated; wherein, the AI ​​resource information recommendation model is obtained by fine-tuning a pre-trained model based on sample data of the target service domain; The first target resource configuration information, the recommendation reason information, and the script information are sent to the first terminal of the target service personnel, so that the target service personnel can recommend the first target resource configuration information to the first user through the communication channel between the first terminal and the second terminal of the first user using the script information and the recommendation reason information.

2. The method according to claim 1, characterized in that, The resource information recommendation request includes the identification information of the first user, the identification information of the first terminal, and the identification information of the second terminal; In response to resource information recommendation requests submitted by target service personnel in the target service domain, the system obtains the initial profile data and in-depth profile data of the first user, including: Based on the identification information of the first user, the user information of the first user is obtained, and the user information includes initial user data and initial service request data; A communication channel is established between the first terminal and the second terminal based on the identification information of the first terminal and the identification information of the second terminal, and the conversation content generated during the conversation between the target service personnel and the first user based on the communication channel is obtained. The conversation content is input into a large language model to perform semantic understanding on the conversation content in order to obtain deep user data and deep service demand data related to the first user. User profiling analysis is performed based on the initial user data and the initial service demand data to obtain the initial profile data of the first user; and User profile analysis is performed based on the deep user data and the deep service demand data to obtain the deep profile data of the first user.

3. The method according to claim 1, characterized in that, Based on the initial profile data and the deep profile data, a potential resource demand analysis is performed on the first user to obtain multi-dimensional potential resource demand information for the first user, including: The initial profile data and the deep profile data are input into the target resource demand analysis model. The initial profile data and the deep profile data are semantically understood from the resource demand dimension to obtain the target resource demand semantic information set. The target resource demand semantic information set includes the resource demand semantic information contained in the initial profile data and the deep profile data, as well as the resource demand semantic information contained in both. Semantic classification is performed on the resource demand semantic information in the target resource demand semantic information set to obtain multiple target semantic categories; Based on the pre-learned mapping relationship between semantic categories and resource demand information in the target service domain, resource demand information is mapped to the multiple target semantic categories to obtain the potential multi-dimensional resource demand information of the first user.

4. The method according to claim 3, characterized in that, Also includes: Obtain a sample dataset in the target service domain. The sample dataset includes multiple sample profile data, sample resource demand semantic information corresponding to each sample profile data, and sample semantic category corresponding to each sample resource demand semantic information. The multiple sample profile data are input into the initial resource demand analysis model, and the multiple sample profile data are semantically understood from the resource demand dimension to obtain intermediate resource demand semantic information. The semantic information of resource demand in the intermediate state is semantically classified to obtain multiple semantic categories of the intermediate state, and the mapping relationship between the semantic categories of the intermediate state and the semantic information of resource demand in the intermediate state is learned. Based on the mapping relationship, the resource requirement information of the semantic categories of the various intermediate states is mapped to obtain the resource requirement information of the intermediate states. Based on the resource demand semantic information of the intermediate state, the sample resource demand semantic information, the semantic category of the intermediate state, and the sample semantic category, the model loss function is calculated. If the model loss function does not meet the model training termination condition, the initial resource demand analysis model is trained until the model loss function meets the model training termination condition, thus obtaining the target resource demand analysis model.

5. The method according to claim 4, characterized in that, The model loss function is calculated based on the resource demand semantic information of the intermediate state, the sample resource demand semantic information, and the semantic category of the intermediate state and the sample semantic category, including: Calculate a first loss function between the resource demand semantic information of the intermediate state and the sample resource demand semantic information, and calculate a second loss function between the semantic category of the intermediate state and the semantic category of the sample, both of which are used as the model loss function; or Calculate a first loss function between the resource demand semantic information of the intermediate state and the sample resource demand semantic information, and calculate a second loss function between the semantic category of the intermediate state and the semantic category of the sample. Then, perform a weighted sum of the first loss function and the second loss function to obtain the model loss function.

6. The method according to claim 1, characterized in that, Each type of resource configuration information has its own applicable conditions; the initial profile data, the deep profile data, the multi-dimensional resource demand information, and the recommendation conditions for various resource configuration information are input into the AI ​​resource information recommendation model. Based on the multi-dimensional resource demand information and the recommendation conditions for the resource configuration information, a first target resource configuration information is selected from the various resource configuration information, including: The multi-dimensional resource demand information, the recommendation conditions for each type of resource configuration information, and the applicable conditions are input into the AI ​​resource information recommendation model. The multi-dimensional resource demand information and the applicable conditions of each type of resource configuration information are semantically matched to determine the target applicable conditions that semantically match the multi-dimensional resource demand information. Resource configuration information that meets the target applicable conditions among the various resource configuration information is selected as candidate resource configuration information; Based on the recommended conditions of the resource configuration information and the multi-dimensional resource demand information, the first target resource configuration information is selected from the candidate resource configuration information.

7. The method according to claim 6, characterized in that, The resource allocation information has multiple recommendation criteria. Based on the resource allocation information recommendation criteria and the multi-dimensional resource demand information, the first target resource allocation information is selected from the candidate resource allocation information, including: Based on the recommendation conditions of each resource configuration information, the various resource configuration information are grouped to obtain an initial set of resource configuration information that matches the recommendation conditions of each resource configuration information. For each initial resource configuration information set, the local adaptability of each resource configuration information in the initial resource configuration information set with the resource demand information of each dimension in the multi-dimensional resource demand information is analyzed from the perspective of service content. The overall adaptability is calculated based on the local adaptability of each resource configuration information with the resource demand information of each dimension in the multi-dimensional resource demand information. The resource configuration information with the overall adaptability greater than the corresponding threshold is used as the first target resource configuration information under the recommendation condition of the resource configuration information.

8. The method according to claim 7, characterized in that, The multi-dimensional resource demand information includes: the first user's attribute information, budget demand level information, and service demand information. The service demand information includes: basic service demand, core service demand, and service demand duration. The recommendation conditions for resource configuration information include optimal resource configuration information recommendation conditions, medium resource configuration information recommendation conditions, and minimum resource configuration information recommendation conditions. The optimal resource configuration information recommendation conditions are used to determine resource configuration information that meets the high budget level and the service demand information. The medium resource configuration information recommendation conditions are used to determine resource configuration information that meets the low budget level and the core service demand and its service demand duration. The minimum resource configuration information recommendation conditions are used to determine resource configuration information that meets the basic service demand and its service demand duration. For each initial resource configuration information set, the local fit degree between each resource configuration information in the initial resource configuration information set and the resource demand information of each dimension in the multi-dimensional resource demand information is analyzed from the service content dimension. The overall fit degree is calculated based on the local fit degree between each resource configuration information and the resource demand information of each dimension in the multi-dimensional resource demand information. The resource configuration information with an overall fit degree greater than the corresponding threshold is used as the first target resource configuration information under the recommendation conditions of the resource configuration information, including: For the initial resource configuration information set under the recommended conditions of the optimal resource configuration information, analyze the first local adaptability of each resource configuration information in the initial resource configuration information set in the dimensions of budget demand level, basic service demand, core service demand, and service demand duration; calculate the first overall adaptability based on the multiple first local adaptability; and take the resource configuration information corresponding to the first overall adaptability being greater than the first threshold as the first target resource configuration information under the recommended conditions of the optimal resource configuration information. For the initial resource configuration information set under the recommended conditions of the medium resource configuration information, analyze the multiple second local adaptability of each resource configuration information in the initial resource configuration information set in the dimensions of budget demand level, core service demand, and service demand duration; calculate the second overall adaptability based on the multiple second local adaptability; and take the resource configuration information corresponding to the second overall adaptability being greater than the second threshold as the first target resource configuration information under the recommended conditions of the medium resource configuration information. For the initial resource configuration information set under the recommended conditions of the guaranteed resource configuration information, analyze the multiple third local fit degrees of each resource configuration information in the initial resource configuration information set in the dimensions of budget demand level, basic service demand, and service demand duration; calculate the third overall fit degree based on the multiple third local fit degrees; and take the resource configuration information corresponding to the third overall fit degree being greater than the third threshold as the first target resource configuration information under the recommended conditions of the guaranteed resource configuration information.

9. The method according to any one of claims 1-8, characterized in that, The recommendation reason information for generating the first target resource configuration information includes: Based on the service requirements pointed to by the multi-dimensional resource requirement information and the configuration services represented by the configuration content of the first target resource configuration information, recommendation reason information for the service requirement class is generated; Based on the budget requirements indicated by the multi-dimensional resource demand information and the configuration price represented by the configuration content of the first target resource configuration information, budget-related recommendation reason information is generated; The recommendation reasons for the service demand category and the recommendation reasons for the budget category are integrated to generate the recommendation reasons for the first target resource configuration information.

10. The method according to claim 9, characterized in that, Also includes: From historical users, a second user is identified. The second user's multi-dimensional resource demand information is more compatible with the first user's multi-dimensional resource demand information than the fourth threshold, and the evaluation level of the second target resource configuration information recommended and purchased by the target service personnel is higher than the fifth threshold. Correspondingly, the recommendation reason information for generating the first target resource configuration information includes: Based on the service requirements pointed to by the multi-dimensional resource requirement information and the configuration services represented by the configuration content of the first target resource configuration information, recommendation reason information for the service requirement class is generated; Based on the budget requirements indicated by the multi-dimensional resource requirement information and the configuration price represented by the configuration content of the first target resource configuration information, budget-related recommendation reasons are generated; based on the second user's multi-dimensional requirement information, the purchased second target resource configuration information, and the historical evaluation information indicating the evaluation level of the purchased second target service, historical evaluation reasons are generated; the service requirement-related recommendation reasons, the budget-related recommendation reasons, and the historical evaluation-related recommendation reasons are integrated to generate the recommendation reasons for the first target resource configuration information.

11. The method according to claim 10, characterized in that, Based on the language style information, generate the script information to recommend the first target resource configuration information to the first user using the recommendation reason information, including: Based on the language style information and the reason information for recommending the first target resource configuration information to the first user, the script information is generated.

12. An electronic device, characterized in that, include: Memory and processor; The memory is used to store a computer program; the processor, coupled to the memory, is used to execute the computer program to implement the steps of the method according to any one of claims 1-11.

13. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it causes the processor to perform the steps of the method according to any one of claims 1-11.

14. A computer program product, characterized in that, The computer program product includes a computer program / instruction that, when executed by a processor, causes the processor to perform the steps of any one of the methods of claims 1-11.

Images