Resource processing method, system and device based on intelligent agent

Abstract

The embodiment of the invention provides an agent-based resource processing method, system and device, and the method comprises the steps: responding to a project request sent by a user client based on a target object, determining a target agent corresponding to the target object, and enabling the target object and the target agent to belong to a target industry dimension; calling the target agent to send data acquisition information associated with the target industry dimension to the user client, and receiving initial resource data returned by the user client based on the data acquisition information; determining target resource data in the initial resource data according to the data acquisition information; and performing project analysis on the target object according to the target resource data to obtain a project request result of the target object. According to the invention, a user can provide digital resource data in a self-service manner, the data acquisition efficiency is improved, and reliable data support is provided for subsequent project analysis.

Description

Technical Field

[0001] The embodiments in this specification relate to the field of artificial intelligence technology, and in particular to a resource processing method, system and apparatus based on intelligent agents. Background Technology

[0002] Credit granting refers to the process by which lending institutions assess a user's comprehensive qualifications, including borrowing behavior, repayment ability, creditworthiness, and loan purpose. After the assessment, the bank provides a preliminary loan amount and the final loan amount based on the results. Currently, for some micro and small enterprises, credit granting before financing relies on in-person visits and assessments by staff, a method that is inefficient and costly. Commonly used video interviews are only available to customers who can authorize online, leaving those unable to access video interviews facing financing difficulties. Therefore, improving the efficiency of credit granting by enabling users to more easily provide offline verification information is a pressing issue. Summary of the Invention

[0003] In view of this, embodiments of this specification provide a resource processing method based on intelligent agents. One or more embodiments of this specification also relate to an intelligent agent-based resource processing system, an intelligent agent-based resource processing device, a computing device, a computer-readable storage medium, and a computer program product, to address the technical deficiencies existing in the prior art.

[0004] According to a first aspect of the embodiments of this specification, a resource processing method based on an intelligent agent is provided, comprising: In response to a project request sent by a user client based on a target object, determine the target intelligent agent corresponding to the target object; The target intelligent agent is invoked to send data collection information associated with the target industry dimension to the user client, and the initial resource data returned by the user client based on the data collection information is received. Based on the data acquisition information, the target resource data is determined according to the initial resource data; Based on the target resource data, the target object is analyzed to obtain the project request results of the target object.

[0005] According to a second aspect of the embodiments of this specification, a resource processing system based on intelligent agents is provided, the system comprising a server and a user client, wherein... The server, in response to the project request sent by the user client based on the target object, determines the target intelligent agent corresponding to the target object, and calls the target intelligent agent to send data collection information related to the target industry dimension to the user client; The user client determines the initial collection data based on the data collection information and sends the initial collection data to the server. The server determines target resource data based on the initial collected data according to the data collection information, performs project analysis on the target object based on the target resource data, and obtains the project request results of the target object.

[0006] According to a third aspect of the embodiments of this specification, a resource processing apparatus based on an intelligent agent is provided, comprising: The response module is configured to respond to a project request sent by a user client based on a target object and determine the target agent corresponding to the target object. The calling module is configured to call the target intelligent agent to send data collection information associated with the target industry dimension to the user client, and receive initial resource data returned by the user client based on the data collection information; The determination module is configured to determine target resource data based on the initial resource data according to the data acquisition information; The analysis module is configured to perform project analysis on the target object based on the target resource data, and obtain the project request results of the target object.

[0007] According to a fourth aspect of the embodiments of this specification, a computing device is provided, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, they implement the steps of the above-described agent-based resource processing method.

[0008] According to a fifth aspect of the embodiments of this specification, a computer-readable storage medium is provided that stores computer-executable instructions, which, when executed by a processor, implement the steps of the above-described agent-based resource processing method.

[0009] According to a sixth aspect of the embodiments of this specification, a computer program product is provided, including a computer program or instructions that, when executed by a processor, implement the steps of the above-described agent-based resource processing method.

[0010] One embodiment of this specification implements the identification of the target intelligent agent corresponding to the target object when a user client sends a project request based on a target object. Both the target intelligent agent and the target object belong to the same target industry dimension. Subsequently, the target intelligent agent can be invoked to send data collection information related to the target industry dimension to the user client. This allows the user client to collect relevant initial resource data based on the data collection information, thereby collecting the data required for the project request. This enables users to provide digitized resource data independently, improving data collection efficiency and providing reliable data support for subsequent project analysis. After obtaining the initial resource data, target resource data can be determined from the initial resource data based on the data collection information, achieving automatic filtering and determination of resource data, reducing manual review costs. Subsequently, project analysis is performed on the target object based on the target resource data to obtain the project request results of the target object. This achieves intelligent and standardized project review and analysis, providing project providers with a convenient and easy project review method, and also providing users with a simple project application method, improving project application efficiency. Attached Figure Description

[0011] Figure 1 A flowchart of an agent-based resource processing method according to an embodiment of this specification is shown; Figure 2 This specification shows a schematic diagram of a resource processing method based on an agent according to an embodiment of the present specification; Figure 3 This specification shows a flowchart illustrating an agent-based resource processing system according to one embodiment. Figure 4 This is a schematic diagram of the structure of a resource processing device based on an intelligent agent, provided in one embodiment of this specification. Figure 5 This is a structural block diagram of a computing device provided in one embodiment of this specification. Detailed Implementation

[0012] Many specific details are set forth in the following description to provide a full understanding of this specification. However, this specification can be implemented in many other ways than those described herein, and those skilled in the art can make similar extensions without departing from the spirit of this specification. Therefore, this specification is not limited to the specific implementations disclosed below.

[0013] The terminology used in one or more embodiments of this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the one or more embodiments of this specification. The singular forms “a,” “described,” and “the” as used in one or more embodiments of this specification and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in one or more embodiments of this specification refers to and includes any or all possible combinations of one or more associated listed items.

[0014] It should be understood that although the terms first, second, etc., may be used to describe various information in one or more embodiments of this specification, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, first may also be referred to as second without departing from the scope of one or more embodiments of this specification, and similarly, second may also be referred to as first. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to a determination."

[0015] Furthermore, it should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in one or more embodiments of this specification are all information and data authorized by the user or fully authorized by all parties. Moreover, the collection, use and processing of related 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.

[0016] First, the terms and concepts used in one or more embodiments of this specification will be explained.

[0017] Artificial Intelligence Agent (AI Agent): An intelligent agent is an intelligent entity capable of perceiving its environment, making decisions, and performing actions. It can take the form of software programs, robots, or virtual assistants.

[0018] AI Real-Time Interaction: AI real-time interaction refers to the process of instant interaction with users or systems using artificial intelligence technology. This interaction can be based on various input forms such as voice, text, and images, and can quickly respond to user requests or behaviors, providing corresponding feedback or services.

[0019] Real-Time Communication (RTC) technology refers to the ability to exchange data with low latency over a network, supporting applications that require instant response, such as voice calls, video conferencing, and online games. RTC technology ensures that information is transmitted almost simultaneously between the sender and receiver, reducing waiting time.

[0020] Large AI models: Large AI models refer to deep learning models with a huge number of parameters and a massive training dataset. These models typically have powerful expressive and generalization capabilities, and can perform complex tasks in a wide range of fields, such as natural language processing and image recognition.

[0021] LUI: Language User Interface (LUI) refers to a language-based user interface. LUI uses natural language processing technology to allow users to communicate with computer systems in a natural language manner. LUI makes human-computer interaction more intuitive and natural, allowing users to complete tasks without needing to learn specific commands or operating procedures.

[0022] Currently, the level of digitalization in the operations of micro and small enterprises is not high, and a large amount of offline business information cannot be digitized and used as the basis for credit granting. This leads to the common pain points of difficulty in financing and insufficient financing amount for micro and small enterprises.

[0023] To address the aforementioned problems, this specification provides a resource processing method based on intelligent agents. This specification also relates to a resource processing system based on intelligent agents, a resource processing device based on intelligent agents, a computing device, a computer-readable storage medium, and a computer program product, which will be described in detail in the following embodiments.

[0024] See Figure 1 , Figure 1 A flowchart of an agent-based resource processing method according to an embodiment of this specification is shown, which specifically includes the following steps.

[0025] Step 102: In response to a project request sent by a user client based on a target object, determine the target intelligent agent corresponding to the target object, wherein the target object and the target intelligent agent belong to the target industry dimension.

[0026] Step 104: Invoke the target intelligent agent to send data collection information associated with the target industry dimension to the user client, and receive the initial resource data returned by the user client based on the data collection information.

[0027] Step 106: Determine the target resource data from the initial resource data based on the data acquisition information.

[0028] Step 108: Perform project analysis on the target object based on the target resource data to obtain the project request results of the target object.

[0029] The user client can be understood as the client used by users who have project application needs. User clients include, but are not limited to, smart devices such as mobile phones and computers. The user client may have a project application provided by the project provider installed. Users can use this application to send project requests based on target objects to the project provider. The target object can be understood as the object of the user's project application. For example, when the project is a loan project, the target object could be the loan recipient within the loan project, such as a residence, shop, or vehicle. Subsequent data collection from the target object is needed as supporting data for the loan project, helping project providers, such as financial institutions, to quickly verify the user's relevant information. To enable users to collect data according to the project application requirements, target intelligent agents belonging to the same industry dimension as the target object can be identified. These target intelligent agents can be understood as AI agents. They can send data collection-related information to the user client according to the interaction process. To flexibly customize the interaction process for different industries, there are also industry-specific distinctions between different intelligent agents. Therefore, in order to collect relevant data from the target object, it is necessary to identify target agents belonging to the same target industry dimension as the target object. This allows the target agents to collect data from the target object according to the data collection information of the associated target industry dimension, enabling the application of the agent-based resource processing method provided in this specification to different industries. After receiving the data collection information, the user client can collect data from the target object based on the data collection information and return initial resource data. The initial resource data is the resource data obtained after collecting data from the target object. The initial resource data can be collected by sensors configured on the user client (such as cameras, microphones, NFC (Near Field Communication)). The initial resource data can include multimodal data such as images, videos, sounds, positioning, and gravity sensing. Subsequently, target resource data can be determined from the initial resource data according to the data collection information, thereby filtering out resource data that matches the data collection information. The target resource data is used to perform project analysis on the target object to obtain the project request results of the target object.

[0030] The following explanation uses a credit application as an example. (See [link to relevant documentation]). Figure 2 , Figure 2 A schematic diagram of an agent-based resource processing method according to an embodiment of this specification is shown. Figure 2In this context, the AI ​​account manager is essentially an AI agent. The AI ​​account manager can determine the appropriate data collection information based on different industry categories, such as... Figure 2 When targeting the catering industry, data collection information can include store location, storefront, business license, kitchen area, and operating area. The AI ​​customer manager, as the front-end interactive entry point of the credit project system, can communicate with users through a LUI interface and guide them through data collection tasks, collecting relevant information customized according to the characteristics of different industries. After data collection, the AI ​​customer manager can also use the large-scale model provided in the backend for real-time risk control analysis, providing service data support for credit services such as loan applications and credit limit increases. Users can upload various types of collected data, such as storefront photos, business license photos, and lease agreements, through user clients such as mobile phones. This collected data forms the basis for analysis and decision-making. The user client can communicate with the credit service provider's server through the ARTC network (Advanced Real-Time Communication). The ARTC network ensures low latency and high reliability of data transmission, supporting real-time interaction in various forms such as voice and video, enabling the AI ​​customer manager to interact with users in real time. The ARTCSDK (Software Development Kit) is a set of development tools to help developers quickly integrate real-time communication functions. Figure 2 The ARTC SDK provides AI account managers with features such as speech-to-text conversion, intelligent noise reduction, and intelligent interruption. It can also return feedback and guidance information generated by the AI ​​account manager through speech synthesis, digital humans, and customized messages to the user client. Correspondingly, to achieve real-time interaction and proactive visual understanding, the AI ​​account manager is configured with various large AI models and services through the model service platform, such as large models, workflow orchestration, and plugin invocation. This provides the AI ​​account manager with powerful computing capabilities and flexible scalability, supporting various application scenarios such as speech synthesis, digital humans, and customized message delivery.

[0031] Based on this, the agent-based resource processing method provided in this specification enables users to conveniently and independently verify information such as offline business premises, operating assets, and business scale. This fills the data gap for credit granting to micro and small enterprises, thereby enabling them to obtain inclusive financing. Simultaneously, it helps financial institutions replace traditional offline manual due diligence verification with intelligent, automated, and online methods, expanding the service boundaries of inclusive finance. The following detailed descriptions of the agent-based resource processing method further illustrate this approach.

[0032] Furthermore, in order to correctly send guidance information for data collection to users, it is necessary to ensure that the industry to which the intelligent agent belongs is the same as the industry of the target object. Therefore, it is necessary to select an intelligent agent with the same industry as the target object as the target intelligent agent. Specifically, determining the target intelligent agent corresponding to the target object includes: obtaining the user attribute information corresponding to the user client, and determining the industry identifier corresponding to the target object based on the user attribute information; and selecting the target intelligent agent corresponding to the target object from the intelligent agent database based on the industry identifier.

[0033] The user attribute information corresponding to the user client can be understood as the attribute information of the user using the user client. This information may include basic user details such as name and phone number, as well as the type of application the user intends to apply for, past application records, the user's industry, and the target entity for this project application. Based on the user attribute information, the industry identifier of the target entity can be determined. For example, if the target entity is a supermarket, the industry identifier could be retail; if the target entity is a restaurant, the industry identifier could be catering. Based on these industry identifiers, an intelligent agent corresponding to the target entity's industry can be selected from the intelligent agent database as the target intelligent agent.

[0034] In practical applications, the AI ​​agent database stores AI agents corresponding to different industries. By classifying industries and pre-setting steps for each, such as the catering and retail industries, expert-level AI account managers are created for each industry. The generated AI agents are stored in the AI ​​agent database. Later, when users have project needs based on target objects, the appropriate AI agent can be selected from the database to provide guidance on data collection.

[0035] In a specific embodiment of this specification, User A is a restaurant owner who wants to apply for a loan from a financial institution to expand his service scope. Lacking standardized financial statements and other traditional credit criteria, User A can use the agent-based credit granting method provided by the financial institution to self-verify his business information. User A submits a loan request through the online self-verification function provided by the financial institution via a corresponding application or mini-program on his mobile phone. After receiving User A's loan request based on his restaurant, the server deployed by the financial institution for this function can obtain the relevant user attribute information pre-filled by User A, learning that User A is engaged in the catering industry and wants a loan based on his restaurant. The user attribute information may also include specific information about the restaurant filled in by the user, such as the type, size, and name of the shop. Based on the identified catering industry identifier, the system searches and matches a target agent in the agent database. The target agent is specifically designed for the catering industry and can be pre-configured with specific data collection task processes or lists for the catering industry, such as taking photos of the storefront, displaying kitchen hygiene conditions, and providing employee health certificates. Subsequently, the target intelligent agent can send data collection guidance information to user A according to the corresponding task process, so that user A can collect data according to the guidance information for subsequent loan review.

[0036] Based on this, by designing AI agents tailored to different industries, expert-level data collection guidance services can be provided for various sectors, facilitating data collection for users and providing reliable data support for subsequent project review. By accurately identifying industry identifiers in user attribute information, it ensures that the appropriate AI agent is selected to meet the specific needs of each industry, improving the professionalism and relevance of the service. AI agents for different industries can be configured with different data collection standards according to their own characteristics, flexibly adapting to diverse market demands.

[0037] Furthermore, in order to collect relevant resource data of the target object, it is necessary to use the relevant hardware sensors configured on the user client. Therefore, it is necessary to obtain the user's permission in advance. Specifically, before calling the target intelligent agent to send the data collection information associated with the target industry dimension to the user client, it also includes: calling the target intelligent agent to send a collection permission request to the user client, receiving permission permission information returned by the user client in response to the collection permission request; establishing a communication channel with the user client based on the permission permission information, wherein the communication channel is used to transmit real-time data sent by the user client.

[0038] In this context, a data collection permission request can be understood as a request for permission to use the data collection device on the user client. This request can be displayed to the user via pop-ups, messages, or other means. The user can grant permission through selection controls within the application interface of the user client. After the server receives the permission permission information returned by the user client, it confirms that the user has agreed to use the relevant data collection device on the user client. This permission permission information represents the user's confirmation of permission to use the data collection device. Based on this permission permission information, a communication channel can be established between the server and the user client. This communication channel is used to transmit real-time data sent by the user client.

[0039] In practical applications, once the server receives the user's permission information, it can activate the relevant data acquisition devices on the user's client, such as cameras and microphones, and access the AI ​​real-time interactive interface. Within this interface, a communication channel is established between the user's client and the server. This communication channel can be understood as an RTC transmission channel, used for the real-time transmission of collected video, audio, and other data.

[0040] In one specific embodiment of this specification, the server sends a data collection permission request to the user client through the application. After the user clicks "confirm," the user client records this operation as permission permission information and sends the permission permission information to the server. A real-time data transmission communication channel is established between the server and the user client, and data is collected and transmitted to the server through the user client's camera, microphone, and other related data collection devices.

[0041] Therefore, before collecting data through the user's client, the system requests the user's permission to collect data. With the user's consent, a real-time transmission channel is established and data is collected and transmitted to protect user privacy and enhance user trust in the project application.

[0042] Furthermore, in order to collect relevant data of the target object, the AI ​​agent will guide the user to collect data according to a preset collection process. Specifically, the AI ​​agent will send data collection information related to the target industry dimension to the user client, including: calling the target AI agent to determine the data collection process related to the target industry dimension, and generating data collection information for the target object based on the data collection process; and sending the data collection information to the user client through the communication channel.

[0043] The data acquisition process can be understood as a set of steps for collecting data specific to a target industry. This process includes the steps for collecting the data required for that industry. For example, if the target industry is the catering industry, the data acquisition process might include steps such as photographing the restaurant's facade, the dining environment, and the kitchen's hygiene. Based on this process, data acquisition information can be generated for the target object. This information can be understood as specific collection requirements and guidance. For instance, if a step in the process involves photographing the restaurant's facade, the corresponding data acquisition information would be "photograph the restaurant facade - including the restaurant name." After understanding this information, users can follow the guidance to collect relevant data, thus providing the required resource data. The server can then send this data acquisition information to the user client via a communication channel. This allows the user client to view the data acquisition information in real-time AI visuals, enabling them to understand the data acquisition process and providing a self-service and convenient way to access data.

[0044] In practical applications, different industry dimensions have different data collection processes. Since AI agents are designed for different industries, their corresponding data collection processes also conform to those industries. The target agent can retrieve the preset data collection process and generate data collection information for the target object based on that process. The server-side AI agent can transmit the collected information to the user client in real time through established communication channels. The user client can then use voice prompts, text prompts, and other methods to demonstrate the current data collection steps in the AI ​​real-time interactive interface.

[0045] In a specific embodiment of this specification, the target intelligent agent is invoked to determine the data collection process related to the catering industry. Based on the data collection process, data collection information for user A's restaurant is generated. The data collection information may include: "1: Please take a clear photo of the storefront, ensuring the signboard is visible; 2: Now please show me your dine-in area, please keep the image stable... Please record a 30-second video of the restaurant's operation." The AI ​​intelligent agent can send the data collection information to the user's client sequentially according to the collection steps. After the user completes the collection task of the current step, it can continue to send the data collection information corresponding to the next collection step to the user's client.

[0046] Based on this, by establishing standardized data collection processes for AI agents in advance for target industry dimensions, we can ensure that the collected resource data is comprehensive and meets risk control requirements. Subsequently, users can effectively avoid omissions or errors in collecting resource data based on the data collection information. Furthermore, users only need to follow the guidance information to complete the collection, reducing the complexity of the data provided by users and enabling them to provide effective project proof data, thereby improving the overall efficiency of the project application process.

[0047] Furthermore, during the process of receiving real-time data transmitted by the user client, if feature data that can be used for data collection is found in the real-time data screen or audio, corresponding guidance information can be generated to guide the user to prioritize or supplement the collection. Specifically, receiving the initial resource data returned by the user client based on the data collection information includes: receiving the initial collection data returned by the user client based on the data collection information, identifying the initial collection data, and obtaining the data content information corresponding to the initial collection data; generating collection guidance information based on the data content information and sending it to the user client; receiving the target collection data returned by the user client based on the collection guidance information, and using the target collection data as the initial resource data.

[0048] The initial data collection can be understood as real-time data transmitted from the user client back to the server. During transmission, the AI ​​agent identifies the initial data collection data, recognizing its content information, which refers to the data characteristics within the initial data collection. For example, if the initial data collection is a picture of a restaurant, and the image contains a restaurant sign, corresponding data collection guidance information can be generated to supplement the data collection for that sign. The user can then retake a close-up picture of the sign based on the guidance information and return this close-up picture as the target data collection data to the server. The server-side AI agent uses the received target data collection data as initial resource data and continues to perform related identification operations.

[0049] In practical applications, during real-time AI interaction, the user client transmits real-time images and other information to the AI ​​agent on the server. The AI ​​agent can use the large AI model configured in the background to recognize and understand multimedia content such as images, videos, sounds, GPS, and gravity sensors, and provide real-time feedback and interaction with the user (through prompts, vibrations, digital humans, etc.).

[0050] In a specific embodiment of this specification, when user A collects relevant resource data of a restaurant based on data collection information, the AI ​​agent identifies the initial collection data returned by the user client. If it detects relatively noisy customer conversations in the audio data of the initial collection data, it analyzes that the restaurant may be in peak business hours. Based on this audio data, it can generate collection guidance information, which is "Please take pictures of the seating arrangement of customers in the restaurant." The collection guidance information is sent to the user client, and the user client prompts the user to collect the corresponding resources according to the collection guidance information through voice broadcast. After user A takes pictures of the dining situation of customers in the restaurant, the collected pictures are transmitted back to the AI ​​agent in real time as target collection data. The AI ​​agent uses them as initial resource data and continues to identify them.

[0051] Based on this, by identifying the initially collected data during real-time data transmission, corresponding collection guidance information can be generated according to the obtained data content information. This information is used to supplement the collection of further usable feature information, improving data integrity and accuracy. This enables AI to "understand" and "comprehend" user actions based on the actual collected content and provide targeted feedback, achieving more natural human-computer interaction. Users are no longer passively performing tasks but receive personalized guidance and assistance, improving user satisfaction. The shift from "user-uploaded data" to "AI-guided collection" makes the data collection process more intelligent and self-service.

[0052] Furthermore, after obtaining the initial resource data, to avoid the initial resource data failing to meet the collection requirements and affecting subsequent project review, it is necessary to verify the initial resource data. Specifically, the target resource data is determined from the initial resource data based on the data collection information, including: determining data verification rules based on the data collection information, and using the data verification rules to verify the initial resource sub-data in the initial resource data; if the initial resource sub-data meets the data verification rules, the initial resource sub-data is used as the target resource data; if the initial resource sub-data does not meet the data verification rules, data re-collection information corresponding to the initial resource sub-data is generated and sent to the user client, and the re-collected sub-data returned by the user client based on the data re-collection information is received; the re-collected sub-data is used as the initial resource sub-data, and the verification of the initial resource sub-data in the initial resource data using the data verification rules continues.

[0053] Data validation rules can be understood as rules used to validate initial resource data. These rules are generated and determined based on data collection information. Data validation rules can be generated in real-time by the AI ​​agent based on the data collection information, or they can be pre-set according to the data collection information. When initial resource data needs to be validated, the corresponding data validation rules are obtained according to the data collection information. For example, if the data collection information is "take a close-up picture of the complete signboard," then the corresponding data validation rule is "detect whether the image contains the complete text of the signboard." After determining the data validation rules, they can be used to validate the initial resource sub-data within the initial resource data. The initial resource sub-data corresponds to the data validation rules. For example, if the data validation rule is to validate "shop signboard," then an image of the "shop signboard" needs to be identified from the initial resource data as the initial resource sub-data, and then validated using the data validation rules.

[0054] In practical applications, after validating the initial resource sub-data using data validation rules, if the initial resource sub-data does not conform to the data validation rules, data re-collection information corresponding to the initial resource sub-data can be generated. This re-collection information can be understood as data re-collected for the content corresponding to the initial resource sub-data. Users can re-collect the relevant content that does not conform to the validation rules based on this information. The re-collected sub-data is then returned to the AI ​​agent, which uses it as initial resource sub-data and re-validates it. If the initial resource sub-data conforms to the data validation rules, it means that the initial resource sub-data meets the collection standards. At this point, the initial resource sub-data can be used as target resource data for subsequent project review.

[0055] In a specific embodiment of this specification, based on the data collection information "Please take a picture of an employee health certificate," the corresponding data verification rule is determined to be "Whether the picture is an employee health certificate." Image resources corresponding to the "employee health certificate" tag are selected from the initial resource data as initial resource sub-data, and verified using the data verification rule. If the verification passes, the initial resource sub-data is used as the target resource data; if the verification fails, data resampling information "Please take another picture of an employee health certificate" is generated. The user can re-collect data based on the data resampling information, and the captured image is returned to the AI ​​agent as re-collected sub-data. The AI ​​agent can then use the data verification rule to verify the user's re-captured image again until it meets the verification rule.

[0056] Based on this, the AI ​​agent automatically verifies the initial resource data according to the data verification rules, avoiding the waste of resources caused by manual review. At the same time, by issuing a re-collection command to the user, the user can re-collect resource data that does not meet the requirements in real time, so that the obtained target resource data has higher clarity and compliance, improving the efficiency and pass rate of subsequent project review.

[0057] Furthermore, to reduce the risk of project application failure due to invalid collected data preventing verification or review during data validation and project review processes, the initial resource data collected can be categorized during real-time data transmission. Specifically, before determining the target resource data from the initial resource data based on the data collection information, the process includes: categorizing the initial resource data according to resource parameters to obtain valid resource data; and determining the target resource data from the initial resource data based on the data collection information, which includes: determining the target resource data from the valid resource data based on the data collection information.

[0058] In this process, classifying initial resource data based on resource parameters can be understood as categorizing the validity of the initial resource data. Resource parameters can be understood as the resolution, clarity, size, and other relevant attributes of each resource data point in the initial data. After classifying the initial resource data using resource parameters, it can be divided into invalid and valid resource data. Invalid resource data can be data that does not meet the acquisition standards, such as blurry or reflective images, unrecognizable audio, or black screen videos. Valid resource data is data that meets the acquisition standards and can proceed to subsequent data verification.

[0059] In practical applications, AI agents can perform pre-processing quality checks on initial resource data transmitted in real time, enabling the identification and classification of image or video content. This allows for rapid judgment and feedback on valid images (black screens, blurriness, reflections, etc.) and valid scenes (street scenes, restaurants, dining areas, etc.). For example, if an AI account manager requests a user to take a photo of a storefront but the user takes a photo of something other than the storefront, or the image is blurry or reflective, the AI ​​will guide the user to adjust their phone to complete the photo. In implementation, after identifying valid resource data, target resource data can be determined from this data. This involves data verification of the valid resource data, filtering out data that can be used as target resource data for subsequent project review.

[0060] In a specific embodiment of this specification, a close-up image of a storefront sign is selected from the initial resource data. The image parameters of this image are obtained, and these parameters are used to classify and identify the image as a resource. If the image parameters do not meet the preset requirements, it is classified as an invalid resource. This method is used to classify the initial resource data and filter out the valid resource data. Subsequently, the target resource data is determined from the valid resource data based on the data collection information.

[0061] Based on this, by automatically classifying and quality-checking the initial resource data transmitted in real time from user clients, valid resource data can be filtered out, ensuring the smooth progress of subsequent review processes. Only valid resource data that meets the standards can enter the subsequent review process. This proactive and dynamic data governance approach not only improves data quality and review efficiency but also enhances the user experience.

[0062] Furthermore, after identifying the target resource data from the initial resource data, project analysis can be performed based on the target resource data. Specifically, project analysis is performed on the target object based on the target resource data to obtain the project request results of the target object, including: inputting the target resource data into an information extraction model to obtain key field information output by the information extraction model; and performing project analysis on the target object based on the key field information to obtain the project request results of the target object.

[0063] The information extraction model can be understood as a large AI model used to extract key information from target resource data. This model can be a multimodal processing model, meaning it can extract key field information from multimodal resource data. Key field information can be understood as the field information containing key content extracted from the target resource data. To facilitate subsequent project review, structured data needs to be extracted from the target resource data. For example, if a piece of resource data is a "close-up image of a shop sign," the extracted key field information could be "shop name = 123"; or if a piece of resource data is an image of seated customers dining in, the extracted key field information could be "number of customers dining in = 20." Subsequent project analysis of the target object can be performed based on the extracted key fields, thereby improving project analysis efficiency. The project request result of the target object is the result obtained after performing project analysis on the target object.

[0064] In practical applications, information extraction models can transform unstructured resource data such as images, videos, and audio into structured fields, providing high-quality input for subsequent modeling and analysis. Specifically, information extraction models utilize methods including, but not limited to, image recognition, OCR recognition, video content understanding, audio recognition, and semantic understanding. These models can automatically extract key information for subsequent project review, reducing the time cost of manual resource data review and improving review efficiency.

[0065] In a specific embodiment of this specification, target resource data is input into an information extraction model to obtain key field information output by the model. This key field information may include phrases such as "store name - 123" and "estimated daily customer traffic - 80 people / day". Based on the extracted key field information, project analysis is performed on the target object to obtain the project request results for that target object.

[0066] Based on this, by introducing an information extraction model, the ability to automatically extract structured field information from target resource data is realized, which greatly shortens the manual review time and improves the overall project review efficiency.

[0067] Furthermore, the project analysis of the target object is performed based on the key field information to obtain the project request result of the target object, including: calculating the project score of the target object based on the key field information; comparing the project score with a preset score threshold; and determining the project request result of the target object based on the comparison result.

[0068] The project score for a target object can be understood as an estimated score calculated based on key fields across various dimensions. This score determines whether the target object can pass the project request. Specifically, project scoring can also be implemented using a large AI model. Key field information is input into the project scoring model, which calculates the target object's project score. Comparing this score with a preset scoring threshold determines the project request outcome. The preset threshold is a pre-defined threshold for the project; if the score is higher than the threshold, the project request is considered passed; if it is lower, the project request is considered rejected.

[0069] In practical applications, the project request results can also include specific result information. For example, if the project is a credit project, the project request results can also include the pre-authorized loan amount.

[0070] In a specific embodiment of this specification, key field information includes "average daily customer traffic: 80 people / day" and "hygiene rating: 4.5 / 5," etc. This key field information is input into the project scoring model. The project scoring model uses multi-dimensional key field information to score the store operated by user A to obtain a project score. The project score is compared with a preset scoring threshold. If the project score is greater than the preset scoring threshold, it is considered approved; if the project score is less than the preset scoring threshold, it is considered unsuccessful or requires the user to provide further information. Correspondingly, an AI model can also analyze the operational status of user A's store and provide credit recommendations, which are then added to the project request results and returned to the user's client.

[0071] This specification provides a resource processing method based on intelligent agents, comprising: responding to a project request sent by a user client based on a target object; determining the target intelligent agent corresponding to the target object, wherein the target object and the target intelligent agent belong to a target industry dimension; invoking the target intelligent agent to send data collection information associated with the target industry dimension to the user client, and receiving initial resource data returned by the user client based on the data collection information; determining target resource data from the initial resource data according to the data collection information; and performing project analysis on the target object based on the target resource data to obtain the project request result of the target object. This method enables the determination of the target intelligent agent corresponding to the target object when the user client sends a project request based on the target object, ensuring that both the target intelligent agent and the target object belong to the same target industry dimension. Subsequently, the target intelligent agent can be invoked to send data collection information associated with the target industry dimension to the user client, allowing the user client to collect relevant initial resource data based on the data collection information, thereby collecting the data required for the project request. This enables users to provide digitized resource data independently, improving data collection efficiency and providing reliable data support for subsequent project analysis. After obtaining the initial resource data, target resource data can be identified from the initial resource data based on the data collection information, enabling automatic filtering and reducing manual review costs. Subsequently, project analysis is performed on the target objects based on the target resource data to obtain the project request results for the target objects. This achieves intelligent and standardized project review and analysis, providing project providers with a convenient and efficient project review method, while also offering users a simple project application method, thus improving project application efficiency.

[0072] Corresponding to the above method embodiments, this specification also provides embodiments of an agent-based resource processing system. Figure 3 A flowchart illustrating an embodiment of an agent-based resource processing system provided in this specification is shown. Figure 3As shown, the system includes a server (302) and a user client (304), wherein, The server 302, in response to the project request sent by the user client based on the target object, determines the target intelligent agent corresponding to the target object, and calls the target intelligent agent to send data collection information related to the target industry dimension to the user client. The user client 304 determines the initial collection data based on the data collection information and sends the initial collection data to the server. The server 302 determines target resource data based on the initial collected data according to the data collection information, performs project analysis on the target object based on the target resource data, and obtains the project request result of the target object.

[0073] In one specific embodiment of this specification, the agent-based resource processing system can be applied not only to credit projects but also to quality inspection projects and factory management projects. That is, the resource processing system can be used to manage the factory's production environment or to inspect the manufactured products.

[0074] This specification provides an agent-based resource management system that, when a user client sends a project request based on a target object, identifies the target agent corresponding to the target object. Both the target agent and the target object belong to the same target industry dimension. Subsequently, the target agent can be invoked to send data collection information related to the target industry dimension to the user client. This allows the user client to collect relevant initial resource data based on the data collection information, thereby acquiring the data required for the project request. This enables users to provide digitized resource data independently, improving data collection efficiency and providing reliable data support for subsequent project analysis. After obtaining the initial resource data, target resource data can be determined from the initial resource data based on the data collection information, achieving automatic filtering and reducing manual review costs. Subsequent project analysis based on the target resource data yields the target object's project request results, achieving intelligent and standardized project review and analysis. This provides project providers with a convenient and efficient project review method, and also offers users a simple project application method, improving project application efficiency.

[0075] Corresponding to the above method embodiments, this specification also provides embodiments of a resource processing device based on intelligent agents. Figure 4 A schematic diagram of the structure of a resource processing device based on an intelligent agent, according to one embodiment of this specification, is shown. Figure 4 As shown, the device includes: Response module 402 is configured to respond to a project request sent by a user client based on a target object and determine the target agent corresponding to the target object; The calling module 404 is configured to call the target intelligent agent to send data collection information associated with the target industry dimension to the user client, and receive initial resource data returned by the user client based on the data collection information; The determination module 406 is configured to determine based on the initial resource data according to the data acquisition information; Analysis module 408 is configured to perform project analysis on the target object based on the target resource data to obtain the project request results of the target object.

[0076] Optionally, the response module 402 is further configured to obtain user attribute information corresponding to the user client, and determine the industry identifier corresponding to the target object based on the user attribute information; and select the target intelligent agent corresponding to the target object in the intelligent agent database based on the industry identifier.

[0077] Optionally, the calling module 404 is further configured to call the target intelligent agent to send a data collection permission request to the user client, receive permission permission information returned by the user client in response to the data collection permission request, and establish a communication channel with the user client based on the permission permission information, wherein the communication channel is used to transmit real-time data sent by the user client.

[0078] Optionally, the calling module 404 is further configured to call the target intelligent agent to determine the data collection process associated with the target industry dimension, and generate data collection information for the target object based on the data collection process; and send the data collection information to the user client through the communication channel.

[0079] Optionally, the calling module 404 is further configured to receive initial collection data returned by the user client based on the data collection information, identify the initial collection data, obtain data content information corresponding to the initial collection data; generate collection guidance information based on the data content information and send it to the user client; receive target collection data returned by the user client based on the collection guidance information, and use the target collection data as the initial resource data.

[0080] Optionally, the determining module 406 is further configured to: determine data verification rules based on the data acquisition information; verify the initial resource sub-data in the initial resource data using the data verification rules; if the initial resource sub-data conforms to the data verification rules, use the initial resource sub-data as target resource data; if the initial resource sub-data does not conform to the data verification rules, generate data re-acquisition information corresponding to the initial resource sub-data and send it to the user client; receive the re-acquisition sub-data returned by the user client based on the data re-acquisition information; use the re-acquisition sub-data as initial resource sub-data; and continue to verify the initial resource sub-data in the initial resource data using the data verification rules.

[0081] Optionally, the determining module 406 is further configured to classify the initial resource data according to resource parameters to obtain valid resource data; and to determine target resource data in the initial resource data according to the data acquisition information, including: determining target resource data in the valid resource data according to the data acquisition information.

[0082] Optionally, the analysis module 408 is further configured to input the target resource data into the information extraction model to obtain key field information output by the information extraction model; and to perform project analysis on the target object based on the key field information to obtain the project request results of the target object.

[0083] Optionally, the analysis module 408 is further configured to calculate the project score of the target object based on the key field information; compare the project score with the preset score threshold; and determine the project request result of the target object based on the comparison result.

[0084] This specification provides a resource processing device based on intelligent agents. When a user client sends a project request based on a target object, it identifies the target intelligent agent corresponding to the target object. Both the target intelligent agent and the target object belong to the same target industry dimension. Subsequently, the target intelligent agent can be invoked to send data collection information related to the target industry dimension to the user client. This allows the user client to collect relevant initial resource data based on the data collection information, thereby acquiring the data required for the project request. This enables users to provide digitized resource data independently, improving data collection efficiency and providing reliable data support for subsequent project analysis. After obtaining the initial resource data, target resource data can be determined from the initial resource data based on the data collection information, achieving automatic filtering and reducing manual review costs. Subsequent project analysis based on the target resource data yields the project request results for the target object, achieving intelligent and standardized project review and analysis. This provides project providers with a convenient and efficient project review method, and also offers users a simple project application method, improving project application efficiency.

[0085] The above is an illustrative scheme of a resource processing device based on intelligent agents according to this embodiment. It should be noted that the technical solution of this resource processing device based on intelligent agents belongs to the same concept as the technical solution of the resource processing method based on intelligent agents described above. For details not described in detail in the technical solution of the resource processing device based on intelligent agents, please refer to the description of the technical solution of the resource processing method based on intelligent agents described above.

[0086] Figure 5 A structural block diagram of a computing device 500 according to one embodiment of this specification is shown. The components of the computing device 500 include, but are not limited to, a memory 510 and a processor 520. The processor 520 is connected to the memory 510 via a bus 530, and a database 550 is used to store data.

[0087] The computing device 500 also includes an access device 540, which enables the computing device 500 to communicate via one or more networks 560. Examples of these networks include Public Switched Telephone Network (PSTN), Local Area Network (LAN), Wide Area Network (WAN), Personal Area Network (PAN), or combinations of communication networks such as the Internet. The access device 540 may include one or more of any type of wired or wireless network interface (e.g., a network interface card (NIC)), such as an IEEE 802.11 Wireless Local Area Network (WLAN) wireless interface, a Wi-MAX (Worldwide Interoperability for Microwave Access) interface, an Ethernet interface, a Universal Serial Bus (USB) interface, a cellular network interface, a Bluetooth interface, or a Near Field Communication (NFC) interface.

[0088] In one embodiment of this specification, the above-described components of the computing device 500 and Figure 5 Other components, not shown, can also be connected to each other, for example, via a bus. It should be understood that... Figure 5 The block diagram of the computing device shown is for illustrative purposes only and is not intended to limit the scope of this specification. Those skilled in the art can add or replace other components as needed.

[0089] The computing device 500 can be any type of stationary or mobile computing device, including mobile computers or mobile computing devices (e.g., tablet computers, personal digital assistants, laptop computers, notebook computers, netbooks, etc.), mobile phones (e.g., smartphones), wearable computing devices (e.g., smartwatches, smart glasses, etc.) or other types of mobile devices, or stationary computing devices such as desktop computers or personal computers (PCs). The computing device 500 can also be a mobile or stationary server.

[0090] The processor 520 is used to execute the following computer-executable instructions, which, when executed by the processor, implement the steps of the above-described agent-based resource processing method.

[0091] The above is an illustrative scheme of a computing device according to this embodiment. It should be noted that the technical solution of this computing device and the technical solution of the above-described agent-based resource processing method belong to the same concept. For details not described in detail in the technical solution of the computing device, please refer to the description of the above-described agent-based resource processing method.

[0092] An embodiment of this specification also provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the above-described agent-based resource processing method.

[0093] The above is an illustrative scheme of a computer-readable storage medium according to this embodiment. It should be noted that the technical solution of this storage medium belongs to the same concept as the technical solution of the above-described agent-based resource processing method. For details not described in detail in the technical solution of the storage medium, please refer to the description of the technical solution of the above-described agent-based resource processing method.

[0094] An embodiment of this specification also provides a computer program product, including a computer program or instructions that, when executed by a processor, implement the steps of the above-described agent-based resource processing method.

[0095] The above is an illustrative scheme of a computer program product according to this embodiment. It should be noted that the technical solution of this computer program product and the technical solution of the above-described agent-based resource processing method belong to the same concept. For details not described in detail in the technical solution of the computer program product, please refer to the description of the above-described agent-based resource processing method.

[0096] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0097] The computer instructions include computer program code, which may be in the form of source code, object code, executable file, or certain intermediate forms. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium may be appropriately added or removed according to the requirements of patent practice. For example, in some regions, according to patent practice, computer-readable media may not include electrical carrier signals and telecommunication signals.

[0098] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the embodiments in this specification are not limited to the described order of actions, because according to the embodiments in this specification, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in this specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the embodiments in this specification.

[0099] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0100] The preferred embodiments disclosed above are merely illustrative of this specification. Optional embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the embodiments described in this specification. These embodiments are selected and specifically described in this specification to better explain the principles and practical applications of the embodiments, thereby enabling those skilled in the art to better understand and utilize this specification.

Claims

1. A resource processing method based on intelligent agents, comprising: In response to a project request sent by a user client based on a target object, the target intelligent agent corresponding to the target object is determined, wherein the target object and the target intelligent agent belong to the target industry dimension; The target intelligent agent is invoked to send data collection information associated with the target industry dimension to the user client, and the initial resource data returned by the user client based on the data collection information is received. Based on the data acquisition information, the target resource data is determined from the initial resource data; Based on the target resource data, the target object is analyzed to obtain the project request results of the target object.

2. The method according to claim 1, wherein determining the target agent corresponding to the target object includes: Obtain the user attribute information corresponding to the user client, and determine the industry identifier corresponding to the target object based on the user attribute information; Based on the industry identifier, select the target intelligent agent corresponding to the target object from the intelligent agent database.

3. The method according to claim 1, before invoking the target intelligent agent to send data collection information associated with the target industry dimension to the user client, further includes: The target intelligent agent is invoked to send a data collection permission request to the user client, and the permission permission information returned by the user client in response to the data collection permission request is received. A communication channel is established with the user client based on the permission information, wherein the communication channel is used to transmit real-time data sent by the user client.

4. The method according to claim 3, wherein the target intelligent agent is invoked to send data collection information associated with the target industry dimension to the user client, comprising: The target intelligent agent is invoked to determine the data collection process associated with the target industry dimension, and data collection information for the target object is generated based on the data collection process; The data collection information is sent to the user client through the communication channel.

5. The method according to claim 1, wherein receiving initial resource data returned by the user client based on the data collection information includes: Receive the initial collection data returned by the user client based on the data collection information, identify the initial collection data, and obtain the data content information corresponding to the initial collection data; Based on the data content information, a collection guidance message is generated and sent to the user client; The system receives the target collection data returned by the user client based on the collection guidance information, and uses the target collection data as the initial resource data.

6. The method according to claim 1, wherein the target resource data is determined from the initial resource data based on the data acquisition information, comprising: Based on the data collection information, data verification rules are determined, and the initial resource sub-data in the initial resource data is verified using the data verification rules. If the initial resource sub-data conforms to the data verification rules, the initial resource sub-data will be used as the target resource data. If the initial resource sub-data does not conform to the data verification rules, generate data re-collection information corresponding to the initial resource sub-data and send it to the user client, and receive the re-collected sub-data returned by the user client based on the data re-collection information; The re-collected sub-data is used as the initial resource sub-data, and the initial resource sub-data in the initial resource data is further validated using the data validation rules.

7. The method according to claim 1, further comprising, before determining the target resource data from the initial resource data based on the data acquisition information: The initial resource data is classified according to the resource parameters to obtain valid resource data; Based on the data acquisition information, target resource data is determined from the initial resource data, including: The target resource data is determined from the valid resource data based on the data collection information.

8. The method according to claim 1, wherein project analysis is performed on the target object based on the target resource data to obtain the project request results of the target object, comprising: The target resource data is input into the information extraction model to obtain the key field information output by the information extraction model. Based on the key field information, the target object is analyzed to obtain the project request results of the target object.

9. The method according to claim 8, wherein project analysis is performed on the target object based on the key field information to obtain the project request result of the target object, comprising: Calculate the project score of the target object based on the key field information; The project score is compared with a preset score threshold, and the project request result of the target object is determined based on the comparison result.

10. A resource processing system based on intelligent agents, the system comprising a server and a user client, wherein, The server, in response to the project request sent by the user client based on the target object, determines the target intelligent agent corresponding to the target object, and calls the target intelligent agent to send data collection information related to the target industry dimension to the user client; The user client determines the initial collection data based on the data collection information and sends the initial collection data to the server. The server determines target resource data based on the initial collected data according to the data collection information, performs project analysis on the target object based on the target resource data, and obtains the project request results of the target object.

11. A resource processing device based on intelligent agents, comprising: The response module is configured to respond to a project request sent by a user client based on a target object and determine the target agent corresponding to the target object. The calling module is configured to call the target intelligent agent to send data collection information associated with the target industry dimension to the user client, and receive initial resource data returned by the user client based on the data collection information; The determination module is configured to determine target resource data based on the initial resource data according to the data acquisition information; The analysis module is configured to perform project analysis on the target object based on the target resource data, and obtain the project request results of the target object.

12. A computing device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions, which, when executed by the processor, implement the steps of the agent-based resource processing method according to any one of claims 1 to 9.

13. A computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the agent-based resource processing method according to any one of claims 1 to 9.

14. A computer program product comprising a computer program or instructions which, when executed by a processor, implement the steps of the agent-based resource processing method according to any one of claims 1 to 9.

Images