E-commerce product compliance detection method and system based on global matching of platform policies

Abstract

The present application provides an e-commerce product compliance detection method and system based on global matching of platform policies. After receiving the product graphic information and detection setting range of the target product uploaded by the user, and receiving the trigger operation of the compliance detection control of the product policy compliance detection interface, a policy compliance detection request message is sent to the first cloud-based intelligent entity; the product sales compliance category is received from the first cloud-based intelligent entity, and the product sales compliance category is detected as a banned product or a restricted product, a root cause detection request message is created, and a root cause detection request message is sent to the first cloud-based intelligent entity; the policy root cause description information is received from the first cloud-based intelligent entity, and the policy root cause description information is displayed on the product policy compliance detection interface. Therefore, the present application can ensure that the target product is detected by each policy related to the product, and output the corresponding violation behavior and policy reasons, thereby improving the comprehensiveness and accuracy of compliance detection.

Description

Technical Field

[0001] The present application relates to the field of information processing technology, and in particular to an e-commerce product compliance detection method and system based on global matching of platform policies. Background Art

[0002] In the process of product compliance testing, existing e-commerce platforms usually use fixed policy terms to conduct compliance checks on listed products. Other illegal or irregular behaviors outside the regular terms cannot be accurately and timely detected, and the efficiency and accuracy of compliance testing are low. Summary of the invention

[0003] The present application provides an e-commerce product compliance detection method and system based on global matching of platform policies, which can ensure that the product will be tested for each policy related to the product and output the corresponding violation behavior and policy reasons, thereby improving the comprehensiveness and accuracy of compliance detection, thereby protecting the interests of sellers and platforms and reducing the legal and economic risks caused by violations.

[0004] In the first aspect, the present application provides an e-commerce product compliance detection method based on global matching of platform policies, which is applied to terminal devices of an e-commerce platform, wherein the e-commerce platform also includes a first cloud-based intelligent agent and a second cloud-based intelligent agent, wherein the first cloud-based intelligent agent and the second cloud-based intelligent agent are deployed on at least one server, and the method includes: receiving product image and text information of a target product uploaded by a user and a detection setting range, wherein the detection setting range includes at least one of a target country and a target platform; and displaying a compliance detection control on a product policy compliance detection interface; and receiving a trigger operation on the compliance detection control, creating a policy compliance detection request message; and sending the policy compliance detection request message to the first cloud-based intelligent agent, wherein the policy compliance detection request message includes the product image and text information and the detection setting range, and the policy compliance detection request message is used to instruct the first cloud-based intelligent agent to generate a first prompt word structure, and the first prompt word structure is used to instruct the second cloud-based intelligent agent of a large language model to determine the product sales compliance category of the target product and send it to the first cloud-based intelligent agent. The first cloud-based intelligent agent outputs the product sales compliance category, which includes banned products, restricted products, and safe products; receives a first detection response message from the first cloud-based intelligent agent, which includes the product sales compliance category; detects that the product sales compliance category is the banned product or the restricted product, creates a root cause detection request message, and sends the root cause detection request message to the first cloud-based intelligent agent, wherein the root cause detection request message is used to instruct the first cloud-based intelligent agent to generate a second prompt word structure, wherein the second prompt word structure is used to instruct the second cloud-based intelligent agent to determine the policy root cause description information of the target product and output the policy root cause description information to the first cloud-based intelligent agent, wherein the policy root cause description information is used to characterize the policy reasons why the target product is banned or restricted; receives a second detection response message from the first cloud-based intelligent agent, wherein the second detection response message includes the policy root cause description information, and displays the policy root cause description information on the product policy compliance detection interface.

[0005] In the second aspect, the e-commerce platform system includes a terminal device, a first cloud-based intelligent agent, and a second cloud-based intelligent agent. The first cloud-based intelligent agent and the second cloud-based intelligent agent are deployed on at least one server, and the terminal device is used to execute the step instructions in the method as described in any one of the first aspects.

[0006] It can be seen that in an embodiment of the present application, product graphic information and a detection setting range of a target product uploaded by a user are received, the detection setting range includes at least one of a target country and a target platform; and a compliance detection control is displayed on the product policy compliance detection interface; and a trigger operation for the compliance detection control is received, a policy compliance detection request message is created; and a policy compliance detection request message is sent to the first cloud-based intelligent entity, the policy compliance detection request message includes the product graphic information and the detection setting range, the policy compliance detection request message is used to instruct the first cloud-based intelligent entity to generate a first prompt word structure, the first prompt word structure is used to instruct the second cloud-based intelligent entity of the large language model to determine the product sales compliance category of the target product and output the product sales compliance category to the first cloud-based intelligent entity, the product sales compliance category includes prohibited products, restricted products, sales products, safety products; receiving a first detection response message from the first cloud-based intelligent body, the first detection response message includes the product sales compliance category; detecting that the product sales compliance category is a banned product or a restricted product, creating a root cause detection request message, and sending a root cause detection request message to the first cloud-based intelligent body, the root cause detection request message is used to instruct the first cloud-based intelligent body to generate a second prompt word structure, the second prompt word structure is used to instruct the second cloud-based intelligent body to determine the policy root cause description information of the target product and output the policy root cause description information to the first cloud-based intelligent body, the policy root cause description information is used to characterize the policy reasons why the target product is banned or restricted; receiving a second detection response message from the first cloud-based intelligent body, the second detection response message includes the policy root cause description information, and displaying the policy root cause description information on the product policy compliance detection interface. Therefore, the present application can ensure that the product will be tested for each policy related to the product, and output the corresponding violation and policy reasons, avoiding the problem of missed detection during policy matching in the prior art, improving the comprehensiveness and accuracy of compliance detection, and thereby protecting the interests of sellers and platforms, and reducing the legal and economic risks caused by violations. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0008] Figure 1 A schematic diagram of the structure of an e-commerce platform provided in an embodiment of the present application;

[0009] Figure 2 A schematic diagram of the structure of another e-commerce platform provided in an embodiment of the present application;

[0010] Figure 3 A flowchart of an e-commerce product compliance detection method based on global matching of platform policies provided in an embodiment of the present application;

[0011] Figure 4 One of the schematic diagrams of the product policy compliance detection interface provided in the embodiment of the present application;

[0012] Figure 5 The second schematic diagram of the product policy compliance detection interface provided in the embodiment of the present application;

[0013] Figure 6 The third schematic diagram of the product policy compliance detection interface provided in the embodiment of the present application;

[0014] Figure 7 A schematic diagram of the structure of a terminal device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0015] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0016] The terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or devices.

[0017] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0018] In the embodiments of the present application, "and / or" describes the association relationship of the associated objects, indicating that three relationships may exist. For example, A and / or B can represent the following three situations: A exists alone; A and B exist at the same time; B exists alone. Among them, A and B can be singular or plural.

[0019] In the embodiment of the present application, the symbol " / " can indicate that the objects associated with each other are in an "or" relationship. In addition, the symbol " / " can also indicate a division sign, that is, performing a division operation. For example, A / B can indicate A divided by B.

[0020] In the embodiments of the present application, "at least one item" or similar expressions refer to any combination of these items, including any combination of single items or plural items, and refer to one or more, and multiple refers to two or more. For example, at least one item of a, b, or c can represent the following seven situations: a, b, c, a and b, a and c, b and c, a, b, and c. Among them, each of a, b, and c can be an element or a set containing one or more elements.

[0021] In the embodiments of the present application, "equal to" can be used in conjunction with greater than, and is applicable to the technical solution adopted when greater than, and can also be used in conjunction with less than, and is applicable to the technical solution adopted when less than. When equal to is used in conjunction with greater than, it is not used in conjunction with less than; when equal to is used in conjunction with less than, it is not used in conjunction with greater than.

[0022] In order to solve the above problems, the present application provides an e-commerce product compliance detection method and system based on global matching of platform policies, which can ensure that the product will be tested for each policy related to the product and output the corresponding violations and policy reasons, thereby improving the comprehensiveness and accuracy of compliance detection, thereby protecting the interests of sellers and platforms and reducing the legal and economic risks caused by violations.

[0023] The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0024] See also Figures 1 to 3 , Figure 1 A schematic diagram of the structure of an e-commerce platform provided in an embodiment of the present application, Figure 2 A schematic diagram of the structure of another e-commerce platform provided in an embodiment of the present application, Figure 3 A flowchart of an e-commerce product compliance detection method based on global matching of platform policies provided in an embodiment of the present application.

[0025] The e-commerce platform 1 includes a terminal device 10 and a first cloud-based intelligent entity 21 and a second cloud-based intelligent entity 22 . The terminal device 10 is communicatively connected with the first cloud-based intelligent entity 21 and the second cloud-based intelligent entity 22 .

[0026] The terminal device 10 may specifically include a front-end device applied to the user side and capable of realizing functions such as data collection and data transmission, and may be a user equipment (UE) such as a mobile phone, a smart phone, a laptop, a digital broadcast receiver, a personal digital assistant (PDA), a tablet computer (PAD), a handheld device, a vehicle-mounted device, a wearable device, a computing device or other processing equipment connected to a wireless modem, a mobile station (MS), a mobile terminal, etc. Alternatively, the terminal device 10 may also be a software application that can be run in the above electronic devices. For example, it may be an APP running on a mobile phone.

[0027] The first cloud agent 21 can realize data transmission, data processing and other backend data processing functions, and the second cloud agent 22 is a large language model formed based on reinforcement learning, which can generate relevant outputs according to the input content. The first cloud agent 21 and the second cloud agent 22 can be deployed on the same server, such as Figure 1 As shown; the first cloud agent 21 and the second cloud agent 22 can also be deployed on two independent servers, such as Figure 2 shown.

[0028] The first server 31 and the second server 32 may be physical servers, or server clusters or distributed systems composed of multiple physical servers. In this embodiment, the number of servers is not specifically limited. Alternatively, they may be cloud servers that provide basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0029] The terminal device 10 is Figure 3 The execution subject of the e-commerce product compliance detection method based on global matching of platform policies is shown, and the e-commerce product compliance detection method based on global matching of platform policies includes the following steps S301-S304:

[0030] In step S301, the terminal device receives the product graphic information and detection setting range of the target product uploaded by the user; and displays the compliance detection control on the product policy compliance detection interface; and receives a trigger operation on the compliance detection control, creates a policy compliance detection request message; and sends the policy compliance detection request message to the first cloud-based intelligent entity.

[0031] The product image information includes marketing images and text description information of the target product to be tested, and the detection setting range includes at least one of a target country and a target platform.

[0032] Among them, the user uploads the product graphic information and sets the detection setting range on the product compliance detection interface. After completing the above operations, clicks the compliance detection control; the terminal device creates a policy compliance detection request message based on the product graphic information and the detection setting range, and sends the policy compliance detection request message to the first cloud intelligent entity.

[0033] Among them, the policy compliance detection request message is used to instruct the first cloud-based intelligent agent to generate a first prompt word structure, and the first prompt word structure is used to instruct the second cloud-based intelligent agent of the large language model to determine the product sales compliance category of the target product and output the product sales compliance category to the first cloud-based intelligent agent. That is, the terminal device sends a policy compliance detection request message to the first cloud-based intelligent agent; after receiving the policy compliance detection request message, the first cloud-based intelligent agent creates a first prompt word structure and inputs the first prompt word structure to the second cloud-based intelligent agent of the large language model; after receiving the first prompt word structure, the second cloud-based intelligent agent obtains the product sales compliance category of the target product based on the product graphic information, the detection setting range, and other relevant information, and outputs the product sales compliance category to the first cloud-based intelligent agent; after receiving the product sales compliance category of the target product, the first cloud-based intelligent agent generates a corresponding first detection response message based on the product sales compliance category, and sends the first detection response message to the terminal device.

[0034] Among them, product sales compliance categories include banned products, restricted products, and safe products. Banned products refer to products that are prohibited from being sold on e-commerce platforms, including but not limited to: contraband: such as guns, ammunition, drugs, flammable and explosive items, etc., dangerous goods: such as toxic chemicals, radioactive substances, etc., counterfeit and inferior goods: such as counterfeit brands, forged origins, substandard quality goods, etc., goods that infringe intellectual property rights: such as pirated books, software, music, film and television works, etc., illegal services: such as unauthorized services or activities, such as gambling, pornographic services, etc. Restricted products refer to products whose sales are restricted, including but not limited to medical products: such as prescription drugs, medical devices, etc., which need to comply with relevant laws and regulations before they can be sold; tobacco products: some countries and regions have strict restrictions and regulations on the sale of tobacco products; alcoholic beverages: some countries and regions have age restrictions and license requirements for the sale of alcoholic beverages; food and health products: need to comply with food safety and quality standards, and some special foods also need to obtain special licenses; cosmetics and personal care products: need to comply with relevant laws and regulations, and some special products also need to obtain special licenses. Safe products refer to other products that are not banned or restricted.

[0035] For specific implementation, see Figure 4 , Figure 4 One of the schematic diagrams of the product policy compliance detection interface provided in the embodiment of the present application, such as Figure 4 As shown, the product policy compliance detection interface 3 includes a product display area 31. Clicking the newly added control in the product display area 31 can display the file upload area 32 and the detection setting area 33. The file upload area 32 is used to upload the product image and text information of the target product, and the detection setting area 33 is used to set the target country and target platform for policy compliance detection. Clicking other products can display the detection results of other products. After the user has uploaded the product image and text information of the target product and set the detection setting range, click the compliance detection control to perform policy compliance detection analysis.

[0036] Step S302: The terminal device receives a first detection response message from the first cloud-based intelligent agent.

[0037] Among them, the first detection response message includes the product sales compliance category.

[0038] When implementing it specifically, Figure 5 As shown, the product policy compliance detection interface 3 also includes a large language model dialogue display box 34. When the first cloud-based intelligent agent generates a first prompt word structure, a first instruction is displayed in the dialogue display box 34. The first instruction is displayed as "What is the product sales compliance category of the target product?". When the second cloud-based intelligent agent generates a product sales compliance category of the target product, the product sales compliance category of the target product is displayed in the dialogue display box 34.

[0039] Step S303: The terminal device detects that the product sales compliance category is the prohibited product or the restricted product, creates a root cause detection request message, and sends the root cause detection request message to the first cloud-based intelligent entity.

[0040] When the terminal device detects that the product sales compliance category of the target product is a prohibited or restricted product, a root cause detection request message is created and the root cause detection request message is sent to the first cloud-based intelligent agent. The root cause detection request message is used to instruct the first cloud-based intelligent agent to generate a second prompt word structure, and the second prompt word structure is used to instruct the second cloud-based intelligent agent to determine the policy root cause description information of the target product and output the policy root cause description information to the first cloud-based intelligent agent; that is, after receiving the root cause detection request message, the first cloud-based intelligent agent generates a second prompt word structure and inputs the second prompt word structure to the second cloud-based intelligent agent of the large language model; after receiving the second prompt word structure, the second cloud-based intelligent agent generates the policy root cause description information of the target product and outputs the policy root cause description information to the first cloud-based intelligent agent; after receiving the policy root cause description information of the target product, the first cloud-based intelligent agent creates a second detection response message based on the policy root cause description information and sends the second detection response message to the terminal device.

[0041] Among them, the policy root cause description information is used to characterize the policy reasons why the target product is banned or restricted from sale; specifically, for banned products, the policy root cause description information includes the reasons why the target product is banned from sale and the specific illegal terms; for restricted products, the policy root cause description information includes the reasons why the target product is restricted from sale, the non-compliant terms, and the specific conditions under which the target product can be listed in compliance with regulations.

[0042] Step S304: The terminal device receives a second detection response message from the first cloud-based intelligent entity and displays the policy root cause description information on the product policy compliance detection interface.

[0043] The second detection response message includes policy root cause description information.

[0044] In specific implementation, if the dialog box 34 shows that the product is prohibited from sale or restricted from sale, see Figure 5 , the first cloud-based intelligent agent generates a second prompt word structure and displays a second instruction in the dialogue display box 34, where the second instruction is "reason for prohibition of sale" or "reason for restricted sale". After the second cloud-based intelligent agent generates the policy root cause description information of the target product, the corresponding reason is displayed in the dialogue display box 34. The product policy compliance detection interface 3 also includes an input box 35, in which the user can input various types of input information such as text, symbols, emoticons, pictures, files, etc., and send it to the dialogue display box 34. The terminal device sends the content input by the user to the second cloud-based intelligent agent for use in the second cloud-based intelligent agent dialogue.

[0045] In some embodiments, the terminal device detects that the product sales compliance category is the safety product, and displays compliance prompt information on the product policy compliance detection interface, where the compliance prompt information is used to indicate that the target product meets the compliance policy of the e-commerce platform.

[0046] For specific implementation, see Figure 6 If the dialog display box 34 displays a safe product, the first cloud-based intelligent agent generates a safety prompt message and displays the safety prompt message on the product policy compliance detection interface 3 .

[0047] It can be seen that in an embodiment of the present application, product graphic information and a detection setting range of a target product uploaded by a user are received, the detection setting range includes at least one of a target country and a target platform; and a compliance detection control is displayed on the product policy compliance detection interface; and a trigger operation for the compliance detection control is received, a policy compliance detection request message is created; and a policy compliance detection request message is sent to the first cloud-based intelligent entity, the policy compliance detection request message includes the product graphic information and the detection setting range, the policy compliance detection request message is used to instruct the first cloud-based intelligent entity to generate a first prompt word structure, the first prompt word structure is used to instruct the second cloud-based intelligent entity of the large language model to determine the product sales compliance category of the target product and output the product sales compliance category to the first cloud-based intelligent entity, the product sales compliance category includes prohibited products, restricted products, sales products, safety products; receiving a first detection response message from the first cloud-based intelligent body, the first detection response message includes the product sales compliance category; detecting that the product sales compliance category is a banned product or a restricted product, creating a root cause detection request message, and sending a root cause detection request message to the first cloud-based intelligent body, the root cause detection request message is used to instruct the first cloud-based intelligent body to generate a second prompt word structure, the second prompt word structure is used to instruct the second cloud-based intelligent body to determine the policy root cause description information of the target product and output the policy root cause description information to the first cloud-based intelligent body, the policy root cause description information is used to characterize the policy reasons why the target product is banned or restricted; receiving a second detection response message from the first cloud-based intelligent body, the second detection response message includes the policy root cause description information, and displaying the policy root cause description information on the product policy compliance detection interface. Therefore, the present application can ensure that the product will be tested for each policy related to the product, and output the corresponding violation and policy reasons, avoiding the problem of missed detection during policy matching in the prior art, improving the comprehensiveness and accuracy of compliance detection, and thereby protecting the interests of sellers and platforms, and reducing the legal and economic risks caused by violations.

[0048] In some embodiments, the process of creating the first prompt word structure by the first cloud agent includes the following steps a to d:

[0049] Step a, the first cloud-based intelligent agent obtains the product graphic information, the detection setting range, and the policy data library, wherein the policy data library includes a conventional basic compliance detection policy sub-library;

[0050] Step b, the first cloud-based intelligent agent generates comprehensive product semantic information based on the product graphic information, wherein the comprehensive product semantic information includes at least one element of product category, product name, marketing keywords, product unit price, after-sales service information, intellectual property information, and sales qualification materials;

[0051] Step c, the first cloud-based intelligent agent determines a target policy sub-library according to the comprehensive semantic information of the product, the detection setting range, and the policy data library, wherein the target policy sub-library includes at least one policy clause, and the at least one policy clause corresponds to at least one element in the comprehensive semantic information of the product;

[0052] In step d, the first cloud-based intelligent agent creates a first prompt word structure according to the comprehensive semantic information of the product and the target policy sub-library.

[0053] Among them, the policy data in the policy data library is collected through code tools, stored in the database, and the product listing policy interface is monitored periodically to obtain updated information regularly and determine the policy type. The policy type includes conventional basic policies and temporary non-standard policies. Conventional basic policies refer to long-term policy regulations issued by the state, including various types of policy clauses such as laws, regulations, and rules. Temporary non-standard policies refer to short-term regulatory clauses temporarily formulated by the state in response to emergencies, such as the price control policy of masks during seasonal epidemics. The difference between the two is that for conventional basic policies, when the target product is subject to compliance testing, the target product is directly determined to be compliant according to the preset mapping relationship; for temporary non-standard policies, when the target product is subject to compliance testing, the target product is determined to be compliant according to the adjusted mapping relationship. The adjusted mapping relationship is obtained by generating an adjustment strategy based on the temporary non-standard policy and then adjusting the mapping relationship.

[0054] Therefore, the policy data library includes a conventional basic compliance detection policy sub-library and may include a temporary non-standard compliance detection policy sub-library. When the policy data library only includes a conventional basic compliance detection policy sub-library, the first cloud-based intelligent agent creates a first prompt word structure according to the above steps a-b.

[0055] Among them, the comprehensive semantic information of the product includes at least one element of the product category, product name, marketing keywords, product unit price, after-sales service information, intellectual property information, and sales qualification materials. The policy data library includes a general basic compliance detection policy sub-library. Each product element corresponds to one or more policy clauses in the general basic compliance detection policy sub-library, and each or more policy clauses are also associated with a type of compliance detection of the second cloud-based intelligent body. For example, see Table 1, which is a corresponding relationship table of general detection clauses corresponding to product elements. Product categories correspond to prohibited product category clauses, prohibited illegal server clauses, etc.; product names and intellectual property information correspond to trademarks and other intellectual property infringement and protection clauses; marketing keywords correspond to policies such as false propaganda and advertising laws; product unit prices correspond to unfair competition, malicious dumping or price gouging, etc.; after-sales service information corresponds to consumer rights protection, infringement liability and other clauses.

[0056] Table 1 Correspondence table of routine testing clauses corresponding to product elements

[0057]

[0058] Among them, for step c "the first cloud-based intelligent agent determines the target policy sub-library based on the product's comprehensive semantic information, the detection setting scope, and the entire policy data library", it specifically includes determining at least one policy clause corresponding to the product element in the product's comprehensive semantic information from the general basic compliance detection policy sub-library based on the target country and the target platform, and the at least one policy clause has a corresponding relationship with the product element.

[0059] In some embodiments, the first prompt word structure includes a first instruction, a first context, first input data, and a first output indication; wherein, the first instruction instructs the second cloud-based intelligent agent to output the product sales compliance category of the target product, the first context includes the target policy sub-library, the first input data includes the comprehensive semantic information of the product, and the first output indication is used to specify that the product sales compliance category output by the second cloud-based intelligent agent is one of the banned product, the restricted product, and the safe product.

[0060] Among them, for step d "the first cloud-based intelligent agent creates a first prompt word structure according to the comprehensive semantic information of the product and the target policy sub-library", it specifically includes: after the first cloud-based intelligent agent receives the policy compliance detection request message, it creates a first instruction for the first prompt word structure; determines the first context of the first prompt word structure according to the target policy sub-library, determines the first input data of the first prompt word structure according to the comprehensive semantic information of the product, and determines the first output indication according to the product sales compliance category.

[0061] It can be seen that in this embodiment, when the policy data library only includes the general basic compliance detection policy sub-library, the first prompt word structure is generated by the first cloud-based intelligent agent, and the first prompt word structure is used as the input of the second cloud-based intelligent agent, so that the second cloud-based intelligent agent searches according to the input information to obtain the product sales compliance category of the target product; since the first prompt word structure includes the target policy sub-library screened out by the first cloud-based intelligent agent, the second cloud-based intelligent agent can directly query and retrieve according to the target policy sub-library, determine the detection items for each product element according to the policy terms in the target policy sub-library, and then determine the relevant information set that the second cloud-based intelligent agent needs to collect according to the detection items, which can reduce the impact of redundant policy data on the query of the second cloud-based intelligent agent, making the retrieval process more efficient.

[0062] In some embodiments, the process of creating the product sales compliance category by the second cloud agent specifically includes the following steps A to D:

[0063] Step A, the second cloud-based intelligent agent obtains the first prompt word structure and the target policy vector data sub-library of the target policy sub-library;

[0064] Step B, the second cloud-based intelligent agent obtains a first set of relevant information based on the product comprehensive semantic information and the target policy sub-library;

[0065] Step C, the second cloud-based intelligent agent vectorizes the product comprehensive semantic information and the first related information set to obtain a product comprehensive semantic vector data set and a first related information vector data set, wherein the product comprehensive semantic vector data set includes vector data of each element in the product comprehensive semantic information;

[0066] Step D: The second cloud-based intelligent agent determines the product sales compliance category of the target product based on each element vector data in the product comprehensive semantic vector data set and the first related information set.

[0067] In some embodiments, step B "the second cloud agent obtains a first set of relevant information according to the product comprehensive semantic information and the target policy sub-library" specifically includes the following steps B1-B6:

[0068] Step B1, the second cloud-based intelligent agent obtains the prohibited / restricted product categories in the target country and the target platform according to the product categories and the target policy sub-library, where the prohibited / restricted product categories are the first category of relevant information subsets;

[0069] Among them, the second cloud-based intelligent agent selects the banned and restricted product categories according to the policy clauses in the target policy sub-library, and uses them as the first category of relevant information subsets. For example, by searching for keywords such as "ban", "restriction", "banned sale", and "restricted sale", the banned / restricted sale product categories are obtained, such as prohibiting the sale of substandard products, prohibiting the sale of certain foods (such as genetically modified foods, expired foods, etc.), restricting and prohibiting the sale of unapproved drugs and medical devices, and restricting the sale and advertising of tobacco products.

[0070] Step B2, the second cloud-based intelligent agent determines the right holder and the scope of authorization and licensing of the target product according to the product name and the intellectual property information, where the right holder and the scope of authorization and licensing are a subset of the second category of relevant information;

[0071] Among them, the second cloud-based intelligent agent uses the National Intellectual Property Office or relevant databases to search for patent, trademark and copyright information of the product according to the product name, and extracts the right holder information related to the product from the search results; and retrieves the license agreement, contract or other legal documents related to the product to understand the specific scope and restrictions of the authorization, including region, time, usage, etc., to obtain the scope of the authorized license.

[0072] Step B3, the second cloud-based intelligent agent obtains a marketing keyword set of other e-commerce platforms according to the target product, and obtains a prohibited promotion keyword set according to the target policy sub-library, and determines a third category of related information subset according to the marketing keyword set and the prohibited promotion keyword set, wherein the third category of related information subset is a prohibited marketing information set;

[0073] Among them, the second cloud-based intelligent agent determines the general purpose and efficacy of the target product based on the marketing keyword set of similar products on other e-commerce platforms. Based on the marketing keyword set, it can determine whether the marketing words of the target product seriously exaggerate the product performance and effect, whether they violate scientific basis, whether there is misleading information or false statements, etc.

[0074] Among them, the second cloud-based intelligent body obtains a set of prohibited publicity keywords based on national and local laws and regulations on advertising and publicity, such as the Anti-Unfair Competition Law and the Advertising Law, and clearly prohibits words or expressions such as "best", "only", "well-known trademarks", etc.

[0075] Therefore, based on the marketing keyword set and the prohibited promotion keyword set, a third category of relevant information subset can be comprehensively obtained.

[0076] Step B4, the second cloud-based intelligent agent obtains a product price set of other e-commerce platforms according to the target product, wherein the product price set is a fourth category of relevant information subset;

[0077] Among them, the second cloud-based intelligent agent obtains the price of the target product on other e-commerce platforms, and based on the prices on other e-commerce platforms, it can determine whether there is any behavior of low-price dumping or price gouging for the target product.

[0078] Step B5, the second cloud-based intelligent agent determines the merchant's legal liability scope and the consumer's legal liability scope according to the target product and the target policy sub-library, wherein the merchant's legal liability scope and the consumer's legal liability scope are the fifth category of relevant information subsets;

[0079] Among them, the second cloud-based intelligent body clarifies the statutory liability scope of merchants and consumers of the target products based on the relevant provisions of the Consumer Protection Law, Civil Code and other relevant provisions in the target policy sub-library. Based on the statutory liability scope of merchants and consumers, it can clarify whether the after-sales liability information of the target products complies with legal provisions.

[0080] Step B6, the second cloud-based intelligent agent obtains standard sales qualification materials according to the target product and the target policy sub-library, wherein the standard sales qualification materials are a subset of the sixth category of relevant information;

[0081] The first related information set includes the first category related information subset, the second category related information subset, the third category related information subset, the fourth category related information subset, the fifth category related information subset and the sixth category related information subset.

[0082] It can be seen that in this embodiment, the second cloud-based intelligent agent determines the relevant information subset corresponding to each product element based on each product element in the product comprehensive semantic information and the target policy sub-library. The relevant information subset is used to perform vector retrieval with the corresponding element vector data to obtain the similarity relationship between each element vector data and the relevant information subset, and the detection result of the policy clause corresponding to each product element is determined based on the similarity relationship.

[0083] In some embodiments, step D "the second cloud-based intelligent agent determines the product sales compliance category of the target product based on each element vector data in the product comprehensive semantic vector data set and the first related information set" specifically includes the following steps D1-D4:

[0084] Step D1, the second cloud-based intelligent agent calculates the vector distance between each element vector data in the product comprehensive semantic vector data set and the corresponding vector data subset in the first related information vector data set;

[0085] Step D2, the second cloud-based intelligent agent obtains each element vector data, each vector data subset in the first related information vector data set, and a mapping relationship set of preset vector distances and size relationships between the two;

[0086] Step D3, the second cloud-based intelligent agent determines the detection result of each element in the comprehensive semantic vector of the product according to each vector distance and the mapping relationship set, and the detection result includes a compliance state, a non-compliance state, and a condition not satisfied state;

[0087] Step D4, the second cloud-based intelligent agent determines the product sales compliance category based on the detection results of each element.

[0088] In step D1, for each product element, the following formula (1) is used to calculate the vector distance d(A, B) between the element vector data A in the product comprehensive semantic vector data set and the vector data B in the corresponding related information subset, and the vector space dimension of the element vector data A and the vector data B is n:

[0089]

[0090] Among them, the mapping relationship set includes the preset vector distance and size relationship between each product element vector and its corresponding related information subset. The detection result of each product element can be determined based on the preset vector distance, the calculated actual vector distance and the size relationship.

[0091] Exemplarily, the product category corresponds to a first preset vector distance, and the size relationship is: the actual vector distance is less than or equal to the first preset vector distance, and the detection result is a non-compliant state; that is, when there is a vector distance less than the first preset vector distance between the product category vector data of the target product and the vector distance of the first type of related information vector data sub-library, it indicates that the product category of the target product is highly similar to the banned / restricted product category, and the product category of the target product belongs to the banned / restricted product category, and the detection result is a non-compliant state.

[0092] The product name and intellectual property information correspond to the second preset vector distance, and the size relationship is: the actual vector distance is greater than or equal to the second preset vector distance, and the detection result is a non-compliant state; that is, when the vector data corresponding to the product name and intellectual property information of the target product and the vector distance of the second category of related information vector data subset is greater than or equal to the second preset vector distance, it indicates that the right holder and the scope of the authorization license of the target product are not within the scope of the authorization license, that is, the target product is an infringing product, and the detection result is a non-compliant state. For example, if the target product is a pirated Black Myth Wukong game, the second category of relevant information subset includes steam, PS5 and other platform entities that are legally authorized by Black Myth Wukong, and the sales entity of the target product is other entities outside the second relevant information subset, it indicates that the entity is illegally selling.

[0093] The marketing keywords correspond to the third preset vector distance, and the size relationship is: the actual vector distance is less than or equal to the third preset vector distance, and the detection result is a non-compliant state; that is, when the vector distance between the marketing keyword vector data and the third category related information vector data subset is less than or equal to the third preset vector distance, it indicates that the marketing keywords of the target product have a high degree of similarity with the third category existing backbone information vector data subset, and the marketing keywords of the target product are keywords for marketing, and the detection result is a non-compliant state.

[0094] The unit price of the product corresponds to the fourth preset vector distance, and the size relationship is: the actual vector distance is greater than or equal to the fourth preset vector distance, and the detection result is a non-compliant state; that is, when the vector distance between the unit price vector data of the product and the fourth category of related information vector data subset is greater than or equal to the fourth preset vector distance, it indicates that the unit price of the product is not within the normal market sales price range, and the detection result is a non-compliant state.

[0095] The after-sales service information corresponds to the fifth preset vector distance, and the size relationship is: the actual vector distance is greater than or equal to the fifth preset vector distance, and the detection result is a non-compliant state; that is, when the vector distance between the after-sales service information vector data and the fifth category of related information vector data subset is greater than or equal to the fifth preset vector distance, it indicates that the merchant's responsibility scope or the consumer's responsibility scope is not within the statutory responsibility scope, and the detection result is a non-compliant state.

[0096] The sales qualification materials correspond to the sixth preset vector distance, and the size relationship is: the actual vector distance is greater than or equal to the sixth preset vector distance, and the detection result is non-compliant; that is, when the vector distance between the sales qualification materials vector data and the sixth category of related information vector data subset is greater than or equal to the sixth preset vector distance, it indicates that the target sales qualification materials are missing or omitted compared to the standard sales qualification materials, and the detection result is that the conditions are not met.

[0097] It can be seen that in this embodiment, according to the corresponding mapping relationship designed for each product element, by clarifying the relationship between the actual vector distance and the preset value, it is possible to quickly identify non-compliant states based on vector distance calculation, thereby reducing the time and cost of manual review.

[0098] In some embodiments, step D4 "the second cloud-based intelligent entity determines the product sales compliance category based on the detection result of each element" includes: when the detection result of each element is the compliance state, the product sales compliance category is the safety product; when there is at least one element whose detection result is the non-compliance state, the product sales compliance category is the banned product; when there is at least one element whose detection result is the condition not met state and there is no non-compliance state, the product sales compliance category is the restricted product.

[0099] It can be seen that in this embodiment, the second cloud-based intelligent agent determines the detection result of each product element of the target product by calculating the vector distance between each product element of the target product and the corresponding subset of relevant information, and determines the product sales compliance category of the target product based on the detection results of multiple product elements of the target product. Since the detection results of multiple product elements are taken into consideration, the comprehensiveness of the detection can be improved; at the same time, based on vector distance analysis, the accuracy of the detection results of a single product element can be improved.

[0100] In some embodiments, the policy data library also includes a temporary non-standard compliance detection policy, and the first context of the first prompt word structure also includes a second related information set, which is an adjustment strategy for the mapping relationship set of the conventional basic compliance detection policy generated according to the temporary non-standard compliance detection policy, and the second related information set is used to instruct the second cloud-based intelligent agent to determine the detection results of the corresponding elements in the comprehensive semantic information of the product based on the second related information set.

[0101] Among them, when the entire policy data database includes conventional basic compliance testing policies and temporary non-standard compliance testing policies, for the temporary non-standard compliance testing policies, the corresponding product elements are determined according to the terms of the temporary non-standard compliance testing policies, and the adjustment strategy for the mapping relationship of the corresponding product elements is determined according to the terms of the temporary non-standard compliance testing policies, and the product elements and the adjustment strategy are determined as the second related information set.

[0102] Exemplary, during the seasonal epidemic period, the mask supervision prohibits the price increase sales policy. For the price detection of conventional products, there is a certain price range for the detection, that is, it can rise or fall within a preset range. According to this clause, the price of masks is prohibited from rising during the seasonal epidemic period, and the detection range of its product unit price is only less than or equal to the unit price of masks. Therefore, during the seasonal epidemic period, for the compliance detection of mask prices, it is necessary to adjust the mapping relationship set originally used in conventional detection, and generate an adjustment strategy for product unit prices according to the policy terms; when the second cloud-based intelligent body receives the first prompt word structure, it detects the existence of a second related information set, and when executing the above steps D1-D4, the mapping relationship of the product unit price is adjusted according to the adjustment strategy. When executing the compliance detection of other products after the current mask product, if the first prompt word structure does not carry an adjustment strategy, the above steps D1-D4 are still executed according to the mapping relationship in the mapping relationship set.

[0103] It can be seen that in this embodiment, when the policy data library also includes temporary non-standard compliance inspection policies, the adjustment strategy of the corresponding product elements is determined according to the temporary non-standard compliance inspection policies, and the second cloud-based intelligent entity makes timely and flexible adjustments to the mapping relationship of the product elements according to the adjustment strategy. At the same time, it can also maintain compliance checks on other conventional products to avoid system confusion.

[0104] In some embodiments, the second prompt word structure includes a second instruction, a second context, second input data, and a second output indication, the second instruction instructs the second cloud-based agent to output the policy root cause description information of the target product, the second context includes the first related information set, the second related information set, and the product comprehensive semantic information, and the output indication is used to specify that the policy root cause description information output by the second cloud-based agent includes the product sales compliance category and the reason for prohibition / restriction on sale.

[0105] Among them, after the first cloud-based intelligent agent receives the root cause detection request message from the terminal device, it generates a second prompt word structure. The creation process of the second prompt word structure is consistent with the creation process of the first prompt word structure, which will not be repeated here.

[0106] In some embodiments, the process of creating the policy root cause description information by the second cloud agent includes the following steps E-H:

[0107] Step E, the second cloud-based intelligent agent obtains the first element of the detection result being the non-compliant state, or the second element of the detection result being the condition not being satisfied state;

[0108] Step F, the second cloud-based intelligent agent determines the target policy terms according to the correspondence between the policy terms in the target policy sub-library and the elements in the product comprehensive semantic information, the first element, and the second element;

[0109] Step G, the second cloud-based intelligent agent generates policy root cause description information of the banned product according to the first factor and the target policy clause; or

[0110] Step H: the second cloud-based intelligent agent generates policy root cause description information of the restricted-sale product based on the second factor and the target policy terms.

[0111] Among them, the second cloud-based intelligent agent obtains the first element whose detection result is a non-compliant state, determines the target policy clause that violates the regulations based on the correspondence between the first element and the policy clauses in the target policy sub-library, and the second cloud-based intelligent agent generates policy root cause description information of the banned product based on the first element and the corresponding target policy clauses. The policy root cause description information at least includes banned products, illegal content and illegal clauses.

[0112] Among them, the second cloud-based intelligent agent obtains the second factor of the detection result that the condition is not met, determines the unmet target policy terms based on the correspondence between the second factor and the policy terms in the target policy sub-library, and generates policy root cause description information of the restricted product based on the second factor and the corresponding target policy terms. The policy root cause description information at least includes the restricted product, unmet terms and content to be supplemented.

[0113] It can be seen that in this embodiment, the second cloud-based intelligent agent automatically generates specific reasons for prohibition or restriction of sales based on the detection results, which can help users identify and correct compliance issues in a timely manner, thereby improving the efficiency and accuracy of compliance management as a whole.

[0114] It should be noted that the specific implementation process of this embodiment can refer to the specific implementation process described in the above method embodiment, which will not be described here.

[0115] With the above Figure 3 For details on the embodiments shown in the drawings, please refer to Figure 7 , Figure 7 A schematic diagram of the structure of a terminal device provided in an embodiment of the present application is shown in FIG. Figure 7 As shown, the terminal device 10 includes a processor 71, a memory 73, a communication interface 72, and one or more programs 731. The one or more programs 731 are stored in the memory 73 and are configured to be executed by the processor 71. The above-mentioned programs include methods for executing the methods described in the above-mentioned embodiments.

[0116] An embodiment of the present application also provides a computer storage medium, wherein the computer storage medium stores a computer program for electronic data exchange, wherein the computer program enables a computer to execute part or all of the steps of any method recorded in the above method embodiments, and the above computer includes an electronic device.

[0117] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0118] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0119] In the several embodiments provided in the present application, it should be understood that the disclosed device can be implemented in other ways. For example, the device embodiments described above are only schematic, such as the division of the above-mentioned units, which is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection of devices or units can be electrical or other forms.

[0120] The units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0121] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0122] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a memory, including a number of instructions to enable a computer device (which can be a personal computer, server or network device, etc.) to execute all or part of the steps of the above-mentioned methods in each embodiment of the present application. The aforementioned memory includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk and other media that can store program codes.

[0123] Those skilled in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable memory, which may include a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0124] The embodiments of the present application are introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method and core idea of ​​the present application. At the same time, for general technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A method for e-commerce product compliance detection based on global matching of platform policies, characterized in that: A terminal device applied to an e-commerce platform, wherein the e-commerce platform further comprises a first cloud-based intelligent agent and a second cloud-based intelligent agent, wherein the first cloud-based intelligent agent and the second cloud-based intelligent agent are deployed on at least one server, and the method comprises: Receive product graphic information and a detection setting range of a target product uploaded by a user, the detection setting range including at least one of a target country and a target platform; and display a compliance detection control on a product policy compliance detection interface; and, upon receiving a trigger operation on the compliance detection control, create a policy compliance detection request message; and send the policy compliance detection request message to the first cloud-based intelligent agent, the policy compliance detection request message including the product graphic information and the detection setting range, the policy compliance detection request message being used to instruct the first cloud-based intelligent agent to generate a first prompt word structure, the first prompt word structure being used to instruct the second cloud-based intelligent agent of the large language model to determine a product sales compliance category of the target product and output the product sales compliance category to the first cloud-based intelligent agent, the product sales compliance category including a prohibited product, a restricted product, and a safe product; receiving a first detection response message from the first cloud-based agent, wherein the first detection response message includes the product sales compliance category; Detecting that the product sales compliance category is the banned product or the restricted product, creating a root cause detection request message, and sending the root cause detection request message to the first cloud-based intelligent agent, the root cause detection request message is used to instruct the first cloud-based intelligent agent to generate a second prompt word structure, the second prompt word structure is used to instruct the second cloud-based intelligent agent to determine the policy root cause description information of the target product and output the policy root cause description information to the first cloud-based intelligent agent, the policy root cause description information is used to characterize the policy reason why the target product is banned or restricted from sale; A second detection response message is received from the first cloud-based intelligent entity, wherein the second detection response message includes the policy root cause description information, and the policy root cause description information is displayed on the product policy compliance detection interface.

2. The method according to claim 1, characterized in that The process of creating the first prompt word structure includes the following steps: Obtaining the product graphic information, the detection setting range, and the policy data library, wherein the policy data library includes a sub-library of general basic compliance detection policies; Generate comprehensive product semantic information based on the product graphic information, wherein the comprehensive product semantic information includes at least one element of product category, product name, marketing keywords, product unit price, after-sales service information, intellectual property information, and sales qualification materials; Determine a target policy sub-library according to the comprehensive semantic information of the product, the detection setting range, and the policy data library, wherein the target policy sub-library includes at least one policy clause, and the at least one policy clause corresponds to at least one element in the comprehensive semantic information of the product; A first prompt word structure is created according to the comprehensive product semantic information and the target policy sub-library.

3. The method according to claim 2, characterized in that The first prompt word structure includes a first instruction, a first context, a first input data, and a first output indication; Among them, the first instruction instructs the second cloud-based intelligent agent to output the product sales compliance category of the target product, the first context includes the target policy sub-library, the first input data includes the comprehensive semantic information of the product, and the first output instruction is used to specify that the product sales compliance category output by the second cloud-based intelligent agent is one of the banned product, the restricted product, and the safe product.

4. The method according to claim 3, characterized in that The process of creating the product sales compliance category specifically includes the following steps: Acquire the first prompt word structure, and acquire the target policy vector data sub-library of the target policy sub-library; Acquire a first relevant information set according to the product comprehensive semantic information and the target policy sub-library; Vectorizing the product comprehensive semantic information and the first related information set to obtain a product comprehensive semantic vector data set and a first related information vector data set, wherein the product comprehensive semantic vector data set includes vector data of each element in the product comprehensive semantic information; The product sales compliance category of the target product is determined based on each element vector data in the product comprehensive semantic vector data set and the first related information set.

5. The method according to claim 4, characterized in that The acquiring of a first set of relevant information according to the product comprehensive semantic information and the target policy sub-library comprises: According to the product category and the target policy sub-library, obtain the prohibited / restricted product category in the target country and the target platform, where the prohibited / restricted product category is a first-category related information subset; Determine the right holder and the scope of authorization and licensing of the target product according to the product name and the intellectual property information, where the right holder and the scope of authorization and licensing are a subset of the second category of relevant information; Acquire a marketing keyword set of other e-commerce platforms according to the target product, and acquire a prohibited promotion keyword set according to the target policy sub-library, and determine a third category of related information subset according to the marketing keyword set and the prohibited promotion keyword set, wherein the third category of related information subset is a prohibited marketing information set; Obtain a product price set on other e-commerce platforms according to the target product, where the product price set is a fourth category of relevant information subset; Determine the merchant's legal liability scope and the consumer's legal liability scope according to the target product and the target policy sub-library, wherein the merchant's legal liability scope and the consumer's legal liability scope are a fifth category of relevant information subsets; Acquire standard sales qualification materials according to the target product and the target policy sub-library, wherein the standard sales qualification materials are a subset of the sixth category of relevant information; The first related information set includes the first category related information subset, the second category related information subset, the third category related information subset, the fourth category related information subset, the fifth category related information subset and the sixth category related information subset.

6. The method according to claim 4, characterized in that The determining the product sales compliance category of the target product according to each element vector data in the product comprehensive semantic vector data set and the first related information set includes: Calculating the vector distance between each element vector data in the product comprehensive semantic vector data set and the corresponding vector data subset in the first related information vector data set; Obtaining each element vector data, each vector data subset in the first related information vector data set, and a mapping relationship set of preset vector distances and size relationships between the two; Determine the detection result of each element in the comprehensive semantic vector of the product according to each vector distance and the mapping relationship set, wherein the detection result includes a compliance state, a non-compliance state, and a condition not satisfied state; The sales compliance category of the product is determined based on the test results of each element.

7. The method according to claim 6, characterized in that The policy data library also includes temporary non-standard compliance detection policies, and the first context of the first prompt word structure also includes a second related information set, which is an adjustment strategy for the mapping relationship set of the conventional basic compliance detection policy generated according to the temporary non-standard compliance detection policy. The second related information set is used to instruct the second cloud-based intelligent agent to determine the detection results of the corresponding elements in the comprehensive semantic information of the product based on the second related information set.

8. The method according to claim 6, characterized in that Determining the product sales compliance category according to the test results of each element includes: When the detection result of each element is the compliance status, the product sales compliance category is the safety product; When the detection result of at least one element is the non-compliant state, the product sales compliance category is the prohibited product; When the detection result of at least one element is that the condition is not satisfied, and the non-compliant state does not exist, the product sales compliance category is the restricted product.

9. The method according to claim 7, characterized in that: The second prompt word structure includes a second instruction, a second context, second input data, and a second output instruction, wherein the second instruction instructs the second cloud-based intelligent agent to output the policy root cause description information of the target product, the second context includes the first related information set, the second related information set, and the product comprehensive semantic information, and the output instruction is used to specify that the policy root cause description information output by the second cloud-based intelligent agent includes the product sales compliance category and the reason for prohibition / restriction of sale; The process of creating the policy root cause description information includes the following steps: Obtaining the first element of the non-compliant state as a result of the test, or the second element of the state of not satisfying the condition as a result of the test; Determine the target policy clause according to the correspondence between the policy clauses in the target policy sub-library and the elements in the comprehensive semantic information of the product, the first element, and the second element; Generate policy root cause description information of the banned product according to the first factor and the target policy clause; or, Generate policy root cause description information of the restricted-sale product based on the second factor and the target policy terms.

10. An e-commerce platform system, characterized in that: The e-commerce platform system includes a terminal device, a first cloud-based intelligent agent, and a second cloud-based intelligent agent. The first cloud-based intelligent agent and the second cloud-based intelligent agent are deployed on at least one server. The terminal device is used to execute the step instructions in the method as described in any one of claims 1-9.

Images