Commodity distribution processing method and device

Abstract

The invention discloses a commodity distribution processing method and device, and relates to the technical field of computers. A specific embodiment of the method comprises the steps of obtaining commodity data of a commodity in response to a transaction related operation executed by a user on the commodity, and obtaining a current delivery address of the user; determining physical environment data corresponding to the current delivery address, and determining a delivery risk assessment result of the commodity based on the commodity data and the physical environment data; and under the condition that the distribution risk assessment result is high risk, displaying a preset risk processing scheme for the transaction related operation, and executing the selected risk processing scheme in response to a selection operation of the user on one of the risk processing schemes. According to the embodiment, the physical environment data is introduced in the early stage of transaction, potential risks such as physical space mismatching can be recognized in advance, a processing scheme is provided for high-risk situations, and therefore the false purchase rate and the return rate of large commodities caused by physical environment limitation are effectively reduced, and the user experience and the platform performance efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a commodity distribution processing method and device. Background Art

[0002] In the current e-commerce landscape, while product size matching technology has played a role in improving the shopping experience, large-ticket items still face high return rates and delivery difficulties due to the mismatch between the product and the physical space. Furthermore, relevant risk warnings are currently scattered throughout customer service consultation and after-sales processes, failing to be deeply integrated with the product purchase process. This prevents effective risk assessment and resolution in the early stages of the transaction, further contributing to high rates of mispurchases and returns. Summary of the Invention

[0003] In view of this, an embodiment of the present invention provides a commodity distribution processing method and device, which can at least solve the problem that the existing technology does not incorporate physical environment data and risk assessment and processing into the e-commerce transaction decision-making process, resulting in high mispurchase and return rates of large commodities due to physical space mismatch.

[0004] To achieve the above-mentioned object, according to one aspect of an embodiment of the present invention, a commodity delivery processing method is provided, comprising:

[0005] In response to a transaction-related operation performed by a user on a product, obtaining product data of the product and obtaining the user's current delivery address;

[0006] Determining physical environment data corresponding to the current delivery address, and determining a delivery risk assessment result of the product based on the product data and the physical environment data;

[0007] In the case where the delivery risk assessment result is high risk, the preset risk handling solutions for the transaction-related operations are displayed, and in response to the user's selection operation of one of the risk handling solutions, the selected risk handling solution is executed.

[0008] To achieve the above-mentioned object, according to another aspect of an embodiment of the present invention, a commodity distribution processing device is provided, comprising:

[0009] an acquisition module, configured to acquire product data of a product and a current delivery address of the user in response to a transaction-related operation performed by the user on the product;

[0010] an assessment module, configured to determine physical environment data corresponding to the current delivery address, and determine a delivery risk assessment result of the product based on the product data and the physical environment data;

[0011] The processing module is used to display the preset risk processing solutions for the transaction-related operations when the delivery risk assessment result is high risk, and execute the selected risk processing solution in response to the user's selection operation of one of the risk processing solutions.

[0012] To achieve the above-mentioned objective, according to another aspect of an embodiment of the present invention, a commodity delivery processing electronic device is provided.

[0013] The electronic device of an embodiment of the present invention includes: one or more processors; a storage device for storing one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors implement any of the above-mentioned commodity delivery processing methods.

[0014] To achieve the above-mentioned purpose, according to another aspect of an embodiment of the present invention, a computer-readable medium is provided, on which a computer program is stored. When the program is executed by a processor, any of the above-mentioned commodity delivery processing methods is implemented.

[0015] To achieve the above objectives, according to another aspect of an embodiment of the present invention, a computer program product is provided. A computer program product according to an embodiment of the present invention includes a computer program that, when executed by a processor, implements the commodity delivery processing method provided in an embodiment of the present invention.

[0016] According to the solution provided by the present invention, one embodiment of the above invention has the following advantages or beneficial effects: by dynamically obtaining product data and the user's current delivery address during the user's operation of the product, and matching the corresponding physical environment data based on the address, an intelligent judgment is made on whether the product data and the physical environment are compatible. This method introduces physical environment data to accurately evaluate the adaptability of products in the early stages of the transaction, can identify potential risks such as spatial mismatch in advance, and provide risk treatment solutions for high-risk situations, and execute the risk treatment solution selected by the user, thereby effectively reducing the mispurchase rate and return rate of large items due to physical environment limitations, and improving user experience and platform fulfillment efficiency.

[0017] The further effects of the above-mentioned non-conventional optional manner will be described below in conjunction with specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The accompanying drawings are provided for a better understanding of the present invention and are not intended to limit the present invention.

[0019] Figure 1 This is a schematic diagram of the main process of a commodity distribution processing method according to an embodiment of the present invention;

[0020] Figure 2It is a schematic diagram of the fusion and credibility weighting of multi-source physical environment data;

[0021] Figure 3 is a flow chart of an optional commodity delivery processing method according to an embodiment of the present invention;

[0022] Figure 4 is a flowchart of another optional commodity delivery processing method according to an embodiment of the present invention;

[0023] Figure 5 is a flowchart of another optional commodity delivery processing method according to an embodiment of the present invention;

[0024] Figure 6 is a flowchart of a specific commodity delivery processing method according to an embodiment of the present invention;

[0025] Figure 7 This is a schematic diagram of the main modules of a commodity distribution processing device according to an embodiment of the present invention;

[0026] Figure 8 is an exemplary system architecture diagram in which embodiments of the present invention may be applied;

[0027] Figure 9 It is a schematic diagram of the structure of a computer system of a mobile device or server suitable for implementing the embodiments of the present invention. DETAILED DESCRIPTION

[0028] The following description of exemplary embodiments of the present invention is made in conjunction with the accompanying drawings, in which various details of the embodiments of the present invention are included to facilitate understanding. These details should be considered as merely exemplary. Therefore, it should be appreciated by those skilled in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0029] It should be noted that in the embodiments of the present application, certain software, components, models and other existing solutions in the industry may be mentioned. They should be regarded as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of the present application, but it does not mean that the applicant has or will necessarily use the solution.

[0030] The embodiments and features of the embodiments of the present invention may be combined unless they conflict. The acquisition, transmission, storage, use, and processing of data in the technical solution of the present invention comply with relevant national laws and regulations, are used for legal and reasonable purposes, are not shared, disclosed, or sold beyond these legal uses, and are subject to supervision and management by regulatory authorities.

[0031] With respect to user information, necessary measures should be taken to prevent unauthorized access to such personal information data, ensure that persons with access to such personal information data comply with relevant laws and regulations, and ensure the security of user personal information. Once such user personal information data is no longer needed, risks should be minimized by restricting or even prohibiting data collection and / or deleting the data. Where applicable, including in certain relevant applications, user privacy should be protected by de-identifying the data, for example, by removing specific identifiers (e.g., date of birth), controlling the amount or specificity of stored data (e.g., collecting location data at the city level rather than at the specific address level), controlling how the data is stored, and / or other methods of de-identification.

[0032] In the current e-commerce landscape, product size matching technology primarily focuses on two approaches: relying on historical user behavior data or relying on manual size input. While these technologies have improved the user shopping experience to a certain extent, large-ticket item transactions still suffer from high return rates and delivery difficulties caused by mismatches between the product and the physical space. Furthermore, manual input is not only cumbersome but also prone to mismatches due to measurement errors or misunderstandings, further exacerbating transaction risks and user experience pain points.

[0033] Specifically, existing solutions generally ignore physical environment data. That is, they are mainly based on the parameters of the product itself and the user's active input, and lack a mechanism for collecting and applying key physical environment data such as door width, elevator configuration, and floor conditions, resulting in a high rate of mispurchase of large items. In addition, when facing cold start scenarios such as new users or new communities, due to the lack of sufficient historical data support, it is difficult for the system to effectively predict potential risks, further exacerbating the possibility of mispurchase. At the same time, the current risk warning mechanism also has a fragmentation problem. Relevant risk warnings are mostly scattered in customer service consultation and after-sales links, and fail to be deeply integrated with the product shopping process, resulting in the inability to effectively intercept high-risk orders in the early stages of the transaction, thereby affecting the overall transaction conversion efficiency and user experience.

[0034] Based on the above background, the core purpose of this solution is to address the high rates of mispurchases and returns caused by mismatches in physical environment data (such as unit door size, elevator configuration, and floor level) during e-commerce transactions for large items. By effectively integrating this key data into the transaction decision-making process, the accuracy of user purchases and the security of platform fulfillment are improved. At the same time, to address the issue of missing data for new users or new communities, we explore the establishment of a multi-level data backup mechanism to achieve reliable environmental prediction and risk assessment during the cold start phase. Furthermore, by embedding intelligent risk assessment and intervention mechanisms at key points in the shopping process (such as browsing, adding to cart, and paying for orders), we promote the shift of risk control from "post-processing" to "pre-emptive prevention," thereby comprehensively improving the scientific nature of user decision-making and the security and controllability of the transaction process.

[0035] See also Figure 1 , which shows a main flow chart of a commodity distribution processing method provided by an embodiment of the present invention, including the following steps:

[0036] S101: In response to a transaction-related operation performed by a user on a product, obtaining product data of the product and obtaining the user's current delivery address;

[0037] S102: Determine physical environment data corresponding to the current delivery address, and determine a delivery risk assessment result of the product based on the product data and the physical environment data;

[0038] S103: When the delivery risk assessment result is high risk, the preset risk handling solutions for the transaction-related operations are displayed, and in response to the user selecting one of the risk handling solutions, the selected risk handling solution is executed.

[0039] In the above implementation, in step S101, the transaction-related operations performed by the user on the product typically include browsing, adding to the shopping cart, and placing an order and paying. This solution adopts corresponding risk assessment methods for different operation stages, but the method for obtaining the physical environment data corresponding to the user's current delivery address remains consistent.

[0040] Product data can be extracted from the product details page, including attribute data (such as specifications) and time series characteristics (such as the return rate of similar products in the past 30 days). Based on these specifications, it can be used to determine whether the product is considered a large item. This solution defines large items as those whose volume or weight exceeds standard logistics standards (e.g., a single side length > 1.5m, weight > 30kg). These items typically require special handling or installation services. This means that only when the user is identified as a large item (such as a refrigerator, bed, mattress, wardrobe, bookcase, etc.) is it necessary to obtain the user's current delivery address.

[0041] During the browsing and adding to shopping cart stages, users usually do not need to manually select a delivery address. At this time, the system will default to the default address in the user's delivery address list, such as Room 1802, Unit 3, Building 36, Community D, Street C, District B, City A. During the order payment stage, users need to manually select a delivery address. For example, when a user buys goods for their parents who live in another place, the delivery address should be the parents' address, not the user's default address. For example, if a user has purchased a new house and is renovating it, the delivery address should be the address of the new house, not its default delivery address. In an optional embodiment, the user may add a new delivery address before browsing or adding to the shopping cart. In this case, the system will use the newly added delivery address as the current delivery address by default to ensure that accurate physical environment data is obtained in subsequent processes.

[0042] After obtaining the user's current delivery address, the system needs to verify its completeness, confirming whether it includes the community, building, unit, and household information (the household number typically includes floor information and specific room information, such as 1802 for the second room on the 18th floor). In actual use cases, users may, out of habit, omit some information, such as the household number, or both the unit and household numbers. In this case, the system can prompt the user to indicate the missing information. The prompt may also include a "Modify" option, which the user can click to complete the missing information in the current delivery address, ensuring that the current delivery address contains the complete community, building, unit, and household information. Only after the user completes the completion process according to the prompts and submits a satisfactory delivery address will the system proceed to "determining the physical environment data corresponding to the current delivery address" to ensure the accuracy and effectiveness of subsequent processes.

[0043] In step S102, this solution invokes a geographic network system (which can be considered a physical environment data storage system, used to store dimensional information such as neighborhoods, buildings, units, and apartment types) and, using a spatial hierarchical modeling approach, retrieves the physical environment data corresponding to the current delivery address from this system, in descending granularity, from household number, unit, building, and neighborhood. This solution can also invoke a home improvement system (which stores data such as product dimensions) to obtain one or more of the product's packaging dimensions and external dimensions.

[0044] The physical environment data includes, but is not limited to, secure delivery access data, and may also include, for example, residential data. Residential data includes at least household type data and building data. Household type data further includes information such as household structure, area, and entrance door dimensions. Building data includes information such as building age, completion date, building type (such as whether there is an elevator), structural features (such as elevator or staircase location, unit door orientation, etc.), and floor height. The entrance door dimension information here can also be used to determine whether there is an order for a door purchased based on the current delivery address. If so, the entrance door dimension information can be obtained based on that order.

[0045] In this solution, "determining the physical environment data corresponding to the current delivery address" can be achieved in a variety of ways, including but not limited to the following three methods: rider measurement data, user-reported data, and inferring physical environment data based on historical transaction records.

[0046] Rider measurement data: A measurement task is generated based on the current delivery address and assigned to a rider. The rider provides on-site service and uses specialized equipment (such as a laser rangefinder) to collect relevant physical environment data. For example, the current delivery address is Unit 502, Building 3, Unit 2, Residential Complex, District C, City A. The rider measures the width of the entrance door to be 90 cm.

[0047] User-reported data: Users actively submit environmental data, such as uploading a photo or manually entering information like door width. This data is verified for authenticity through image recognition (such as OCR) or manual review. For example, a user uploaded a photo of an entrance door and manually entered the door width as 87 cm. The system then verified the width through image recognition as 86 cm.

[0048] Inferring physical environment data based on historical transaction records: Based on the size of large items that have been successfully delivered to the same or nearby delivery address, inferring safe delivery access data for that address. Historical transaction records here specifically refer to successful transactions that were not returned due to size issues, covering information such as the same residential complex, building, unit, or apartment type.

[0049] It's important to note that the safe delivery access data in Method 3 isn't directly measured, but rather an indicator of environmental adaptability derived from the actual results of whether goods can be delivered smoothly to households. This serves as indirect but highly valuable evidence of accessibility. For example, if a refrigerator was successfully purchased at an address within the past three months, and the maximum diagonal length of the elevator at that address is 84cm, it can be inferred that the address supports the passage of items smaller than or equal to 84cm.

[0050] In summary, the first two methods provide direct measurement data, while method three uses historical transaction records to reflect actual traffic conditions, serving as an effective indirect basis for environmental adaptability. These three methods can be used independently or in combination to enhance the comprehensiveness and accuracy of the assessment results. Furthermore, based on the physical environment data obtained by the first two methods, the knowledge graph of the neighborhood where the current delivery address is located can be updated.

[0051] This solution can use multi-source data fusion and credibility weighting mechanism to obtain comprehensive physical environment data when comprehensively considering the physical environment data obtained by various implementation methods. Figure 2 As shown in the figure, confidence weights are set according to different data sources. For example, the confidence level for rider-measured data is 0.9, the confidence level for user-reported data is 0.6, and historical transaction records are assigned different confidence levels based on a time decay factor (e.g., 0.8 for the past three months, 0.6 for 3-12 months, and 0.3 for over a year). The final expression for safe delivery access data is: Environment_Score = (w1×D1+w2×D2+w3×D3) / (D1+D2+D3), where wi represents the safe delivery access data obtained from each channel, and Di represents the confidence level.

[0052] For example, let's assume that based on the product specifications, the maximum diagonal length of the product is 85 cm. The safe delivery data measured by the delivery driver is 90 cm, the user reported data of 86 cm, and the safe delivery data inferred from historical orders is 84 cm (assuming the order was placed within the past three months, the weight is 0.8). The calculation result is: Environment_Score = [(0.9×90)+(0.6×86)+(0.8×84)] / [0.9+0.6+0.8] = 199.8 / 2.3≈86.87 cm. This calculation indicates that the address can accommodate products up to approximately 86.87 cm in size. Since 85 cm is less than 86.87 cm, the refrigerator can be safely delivered to the home.

[0053] By integrating multi-source data and a credibility-weighted mechanism, the system improves the accuracy and reliability of product-environment compatibility assessments. Throughout the calculation process, data dimensions are standardized to a common reference standard. For example, all data is measured in centimeters (cm), and the maximum passable width or diagonal length is used as the evaluation basis to ensure consistent evaluation criteria.

[0054] This solution configures corresponding risk assessment methods for different operation stages. Therefore, it can process product data and physical environment data through risk assessment methods corresponding to transaction-related operations to obtain distribution risk assessment results, thereby realizing risk assessment of the entire process.

[0055] In step S103, if the delivery risk assessment result is low risk, no risk treatment is performed. If the delivery risk assessment result is high risk, it is necessary to display the risk treatment plan preset for the transaction-related operations, such as one or more of the following: warning prompt, AR simulation placement, appointment measurement, product modification, and continued order placement.

[0056] If the current transaction-related operation is a browsing operation and the delivery risk assessment result is high risk, a real-time alert function can be configured on the product details page. The alert methods include but are not limited to the following three examples:

[0057] Display warning information below the main product image, such as a red warning bar, to remind users that there may be a risk of physical space mismatch, such as "Because the delivery address you filled in has no elevator and the corridor is narrow, this large furniture may have difficulty moving into the house"; if the user clicks on the information, the decision-making process can also be displayed to inform the user why the space does not match.

[0058] An "AR (Augmented Reality) Simulated Placement" option is provided. In response to a user clicking on the AR simulated placement option, AR images are generated and displayed based on the product data and physical environment data of the product, respectively, to assist the user in intuitively determining whether the product is compatible with the physical space, such as performing simulated placement operations based on the generated AR images.

[0059] A free measurement appointment option is provided. After the user selects this option, the system automatically generates a measurement order based on the current delivery address and assigns it to the rider to receive the physical environment data uploaded by the rider, thereby improving decision-making accuracy and user experience.

[0060] If the delivery risk assessment result is high when the current transaction is an add-to-cart operation, a corresponding alert mechanism will be triggered, such as a pop-up window to clearly inform the user of the current risk. If the user clicks on this information, the decision-making process can also be displayed to inform the user why the space is not compatible.

[0061] If the delivery risk assessment results in a high risk, a prompt window will pop up, providing a detailed risk description and various possible risk mitigation options. Users must select a solution in this window before submitting their order, ensuring they make a fully informed decision. These options include, but are not limited to, modifying the product, scheduling a measurement service, and confirming the risk before continuing with the order.

[0062] If the user chooses to modify a product, the system automatically recommends compatible products in the same category that match the current physical environment data based on the product's category, and displays them in a list for the user to choose from. Once the user selects a product from the recommended list, the system automatically completes the product replacement operation, achieving "zero-perception order change." This mechanism completely preserves the original order context, including the order number, delivery address, delivery method, and coupon usage (such as re-matching discounts that meet the minimum purchase requirement), without interrupting the user's operation process, ensuring that user rights are not affected.

[0063] For example, when a user adds a refrigerator to their shopping cart and proceeds to checkout, the system prompts them that the refrigerator is too large, posing a delivery or delivery risk. The user selects "Modify Product," and the system recommends several compatible refrigerator models based on physical environment data. After the user selects a refrigerator, the system automatically updates the product information while preserving contextual information such as the original order number, shipping address, coupons, and shipping method. The user can then proceed with the payment process without having to resubmit information or adjust configurations, achieving a seamless experience for the user.

[0064] If the user chooses the appointment measurement option, the system will automatically generate a measurement order based on the current delivery address and dispatch a rider to collect the actual physical environment data of the changed address to obtain more accurate information for subsequent evaluation and decision support.

[0065] If the user chooses to proceed with the order after fully understanding the risks, they will be required to sign an enhanced risk disclosure agreement. After confirming the user's signature, the system will allow them to proceed with the order and mark the resulting order as high-risk. The order will then be submitted to the customer service system for specialized fulfillment tracking and service assurance. At the same time, the system will generate a complex work order, including measurement tasks and delivery arrangements, based on the risk type and the user's selected actions. These orders will be uniformly assigned to drivers, reducing repeated visits and improving fulfillment efficiency and user satisfaction.

[0066] Marked fields include, but are not limited to, risk type (e.g., insufficient aisle dimensions, floor restrictions, etc.), comparison of maximum product dimensions to environmental aisles, safety margin setting rationale, risk probability output by the risk assessment model, and whether the user has signed the enhanced disclosure agreement. Customer service personnel can view complete physical environment data, risk analysis process, and key parameters in the system, providing comprehensive data support and decision-making basis for subsequent manual communication, contract performance assessment, and customer service.

[0067] Furthermore, during the order and payment phase, not only is the product's physical environment data re-verified to confirm its compatibility with the current delivery address, but the system can also check for key changes: whether the user has changed the delivery address, modified the delivery method (such as whether disassembly and relocation services are allowed), and whether the system has received new measurement data or the order status has changed. This final, comprehensive verification ensures that even if the user adjusts the delivery address or delivery configuration multiple times during the purchase process, the system can still provide accurate risk warnings based on the latest information, effectively preventing accidental purchases and subsequent returns.

[0068] The method provided in the above embodiment dynamically obtains product data and the user's current delivery address during the user's product operation, and performs a delivery risk assessment based on the physical environment data corresponding to the address, thereby achieving intelligent judgment on whether the product is compatible with the physical environment. This method introduces physical environment data to accurately assess product compatibility in the early stages of the transaction, can identify potential risks such as spatial mismatch in advance, and provide risk management solutions for high-risk situations, executing the risk management solution selected by the user, thereby effectively avoiding the problem of mispurchase of large items due to physical environment limitations, reducing the mispurchase rate and return rate of goods, and improving user experience and platform fulfillment efficiency.

[0069] See also Figure 3 , shows a flow chart of an optional commodity delivery processing method according to an embodiment of the present invention, including the following steps:

[0070] S301: Determine the category to which the product belongs, and obtain a preset safety margin for the category;

[0071] S302: Acquire dimension data from the commodity data, and calculate a cumulative value of the dimension data and the safety margin;

[0072] S303: In response to the accumulated value being greater than the safe delivery passage data in the physical environment data, determining that the delivery risk assessment result is high risk;

[0073] S304: extracting spatial features from the physical environment data; extracting product attributes from the product data; and generating cross-features based on the product attributes and spatial features;

[0074] S305: Determine the category to which the product belongs, and obtain the return rate of products in the same category within a preset period of time;

[0075] S306: Inputting the spatial features, the product attributes, the cross features, and the return rate into a preset risk assessment model to obtain a risk probability value;

[0076] S307: In response to the risk probability value being greater than or equal to a preset risk probability threshold, determining that the delivery risk assessment result is high risk.

[0077] To accurately determine the suitability of different product categories, this solution introduces a dual-path evaluation system: a "rules engine + machine learning model (or risk assessment model)." This system, combined with a multi-task learning framework for classification optimization, ensures efficient judgment while enhancing adaptability and generalization to complex scenarios. The machine learning model's input features include spatial features, product attributes, temporal features, and cross-features, generating a more refined risk probability value through these multi-dimensional features. This model utilizes a multi-task learning framework, extracting common features from a shared base layer to address the suitability of different products. The model training process is similar to the testing process, differing in that risk probability values ​​are annotated during training.

[0078] The above-mentioned rule engine is used to quickly determine the risk conditions for passage. This solution sets up category-differentiated safety margin rules. Different categories of goods face different challenges during transportation and installation. For example, refrigerators are mainly limited by the width of the entrance door and the elevator passage, while air conditioner outdoor units are more dependent on outdoor installation space and hoisting conditions. Therefore, when evaluating the compatibility of the physical environment data of the current delivery address with the product, the system will dynamically determine the safety margin based on the category to which the product belongs. The safety margin here refers to the additional size space reserved in the space planning of elevators, entrance doors, etc. when delivering goods to prevent insufficient physical space or risk of cargo damage due to measurement errors, packaging deformation, or transportation vibration. See the safety margin example shown in Table 1. Therefore, the rule engine is used to determine whether the product can be delivered safely, that is, to determine whether the sum of the product size data and the safety margin is less than or equal to the safe delivery passage data.

[0079]

[0080] When a user browses a product detail page, after obtaining safe delivery data, the system identifies the product's category and retrieves the preset safety margin for that category. It then extracts the actual dimensions from the product data and calculates the cumulative value of the actual dimensions and the safety margin. If this cumulative value is greater than the safe delivery data, a delivery obstacle is determined and the delivery risk assessment is high. Otherwise, it is marked as low risk.

[0081] Assume that the current delivery address is a residence in a newly delivered community A, with a three-bedroom, two-living room apartment, but the measured physical environment data has not yet been collected. The user plans to purchase a refrigerator with a diagonal size of 90 cm. Through intelligent matching, the system finds a mature community B with the same developer and the same apartment type within a 1km radius of community A and retrieves its physical environment data: the entrance door is 85cm wide, the elevator diagonal is 105cm, the floor is 6, and it is a staircase building, as a reference. This data can come from the historical rider measurement data, user-reported data, and historical order data of community B, and there are no restrictions here.

[0082] The system conducts an assessment based on the physical environment data of Community B: While the product's 90cm size is smaller than the 105cm diagonal of the elevator, meeting basic access requirements, the actual required space is 110cm, considering the 20cm safety margin required for refrigerators, exceeding the 105cm diagonal limit for the elevator. Therefore, the system determines that the user's purchase of this product presents a high risk. Furthermore, if the user selects an outdoor air conditioner unit (which is similar in size but requires consideration for outdoor installation), the system will further determine whether there is sufficient installation space. If no suitable installation area is detected, the product will still be marked as high risk, even if it can be successfully brought indoors. This approach ensures risk coverage throughout the entire process, from entrance to final installation.

[0083] When a user browses a product, if the delivery risk assessment result is low, the product information will be displayed normally. When the user adds the product to the shopping cart, the system triggers a complete risk assessment process again. First, the user's latest delivery address is obtained. This address can be the default delivery address or a newly added address before the user adds the product to the shopping cart.

[0084] Since users must have browsed an item before adding it to their shopping cart, and this solution already acquired the delivery address and its corresponding physical environment data during the product browsing phase, it can determine at this stage whether the user changed the default delivery address or added a new one between browsing and adding the item to the shopping cart. If no change occurred, the physical environment data determined by the user when browsing the item will be used. If an address change is detected, the corresponding physical environment data will be re-acquired based on the latest delivery address determined when adding the item to the shopping cart.

[0085] After obtaining accurate physical environment data, the system extracts key spatial features from the residential data, such as building type (elevator / staircase), floor height, building age, entrance door dimensions, and elevator dimensions. It also extracts product attributes and temporal features from the product data. Product attributes include aspect ratio, volume, and weight, while temporal features include indicators such as the return rate of similar products within a certain period of time (e.g., the last 30 days). Furthermore, this evaluation incorporates cross-features, such as the relationship between product length and width and door frame ratio, product weight and floor height, and the degree of matching between product volume and elevator capacity. These cross-features significantly enhance the model's adaptability and prediction accuracy for complex delivery and installation scenarios.

[0086] Regarding the ratio of product length and width to doorframe, if a product is 80cm wide and the doorframe is 85cm wide, the product can be passed through the doorframe at an angle. However, if the product is wider than the doorframe (e.g., a 90cm product and an 85cm doorframe), it may not fit into the house. It is generally recommended that the maximum product width does not exceed 90% of the doorframe to ensure smooth passage.

[0087] Regarding the relationship between product weight and floor height, for example, if a 30kg item needs to be delivered to the 6th floor and there's no elevator, manual handling may be required, potentially affecting delivery methods or increasing shipping costs. Considering delivery difficulty and safety risks, restrictions are often set for the "high floor + heavy item" combination, or elevator transportation is recommended.

[0088] Regarding the compatibility of product size with elevator capacity, if the elevator interior space is 150cm x 120cm and the product is 80cm x 50cm x 40cm, the product will fit easily into the elevator. If the diagonal length of the product is smaller than the elevator door dimensions, you can try to enter diagonally. However, if the product is too large (e.g., its side length exceeds the elevator door width), it will not fit into the elevator even if it is not fully loaded.

[0089] These characteristics are then fed into the risk assessment model to calculate a risk probability value. If this value is greater than or equal to the preset risk probability threshold, the delivery risk assessment is considered high, triggering a corresponding alert mechanism, such as a pop-up window, to clearly inform the user of the current risk.

[0090] If the delivery risk assessment result is low during the product browsing phase, or low during the add-to-cart phase, the user enters the payment phase. During this phase, the system first determines whether the delivery address used for order payment is consistent with the delivery address used during the previous transaction (i.e., the product browsing phase or the add-to-cart phase). If so, the physical environment data determined during the previous transaction is used. If an address change is detected, the corresponding physical environment data is re-acquired based on the latest delivery address determined during the order payment phase.

[0091] Assuming the previous transaction-related operation was a browsing operation, and the user's current delivery address has not changed, then in this case, the delivery risk assessment operation can be omitted and the delivery risk assessment results from the browsing phase can be directly used. Alternatively, the delivery risk assessment operation can be re-evaluated, for example, using a different risk assessment method. Assuming the previous transaction-related operation was an add-to-cart operation, and the user's current delivery address has not changed, then in this case, the delivery risk assessment operation can be omitted and the delivery risk assessment results from the browsing phase can be directly used. Alternatively, a different risk assessment method can be used to re-evaluate.

[0092] However, if the user's current delivery address changes, a new delivery risk assessment is required to re-verify the compatibility of the product and the physical environment data, such as identifying the safety margin corresponding to the category to which the product belongs; extracting the actual size data from the product data, and calculating the cumulative value of the actual size data and the safety margin; if the cumulative value is greater than the safe delivery pass data, it is determined that there is a pass obstacle and the delivery risk assessment result is high risk.

[0093] If the delivery risk assessment result is high risk, a prompt window will be popped up for the high risk situation, which includes a detailed risk description and a variety of optional risk handling solutions. Users must complete the selection of solutions in the window before submitting the order, ensuring that users make the final decision with full knowledge. Optional risk handling solutions include but are not limited to modifying the product, making an appointment for measurement service, confirming the risk and continuing to place the order. For details, please refer to Figure 1 Description shown.

[0094] The method provided in the above embodiment sets up a progressive risk assessment mechanism and deeply integrates it into the entire process of e-commerce transactions, aiming to achieve dynamic early warning capabilities in the entire process from product browsing-adding to shopping cart-payment and ordering, or product browsing-payment and ordering. At the same time, intelligent prompts and intervention measures are set to form a closed-loop control. At each stage, the latest and most complete physical environment data is obtained in combination with the current context to ensure that risk judgments are always close to the real scene. This method not only improves the system's fault tolerance and assessment accuracy, but also significantly optimizes the consistency of user operations and the smoothness of the experience, thereby effectively reducing the risk of misjudgment due to address changes, untimely information updates, or missing data.

[0095] See also Figure 4 , shows a schematic flow chart of another optional commodity delivery processing method according to an embodiment of the present invention, including the following steps:

[0096] S401: Acquire residential data corresponding to the current delivery address from a knowledge graph corresponding to the community information; wherein the residential data includes current building data corresponding to the building information, and current apartment type data corresponding to the building information, unit information, and apartment number information;

[0097] S402: Obtain the safe delivery access data corresponding to the residential data based on the historical transaction records of the product in the order of household number, unit, building, and community granularity from fine to coarse.

[0098] Each community is designed with a knowledge graph to store building data, apartment data, etc. This solution has built a four-level data backup mechanism to ensure that the acquired product data has a high degree of accuracy:

[0099] First, we use historical transaction records for the same apartment type as a reference. These records refer to transactions where installations have been successfully completed without returns. These records contain key data, such as the diagonal dimensions of the elevator. This is achieved by filtering out historical transaction records for the same residential complex, building, unit, and apartment type from the product's historical transaction records. This data is then used to extract safe delivery access data. For example, the maximum pass-through width for a refrigerator in a 3-bedroom, 2-living room apartment is used to assess whether the refrigerator currently being viewed is accessible.

[0100] If no historical transaction records are found for the same residential complex, building, unit, and apartment type, it means that other users in the unit with the same apartment type as the current delivery address have never purchased the product. In this case, the reference range will be expanded to the entire unit. Specifically, the system will retrieve the historical transaction orders for the product in the same unit to extract safe delivery access data from them. The specific implementation method is: from the historical transaction record set of the product, filter out the historical transaction records that meet the conditions of the same residential complex, building, and unit, and extract the corresponding safe delivery access data from them as a reference for whether the product is accessible at the user's current delivery address. In this way, in the absence of data on the same apartment type, it can still provide more accurate risk assessment and logistics judgment support.

[0101] If no historical transaction records are found for the same residential complex, building, or unit, it means that no user in that unit has ever purchased the product. The system then expands the reference scope to include the entire building. Specifically, the system retrieves the secure delivery access data from the historical transaction records for the product within the same building. This is accomplished by filtering out the first historical transaction record for the product within the same residential complex and building, and then obtaining the secure delivery access data from that first historical transaction record.

[0102] As a preferred embodiment, the system can further filter the first historical transaction record selected above based on apartment type similarity to improve the accuracy and applicability of data matching. This is achieved by obtaining other delivery addresses from the first historical transaction record selected, and calculating the apartment type similarity (e.g., area similarity) between the current apartment type data and the apartment type data of the other delivery addresses. Here, apartment type similarity can be calculated as 1-|(difference between the current apartment type data and the apartment type data of the other delivery addresses) / current apartment type data|. The system then selects a second historical transaction record with an apartment type similarity greater than or equal to a preset apartment type similarity threshold and extracts safe delivery access data from it.

[0103] As another preferred implementation, during the process of obtaining safe delivery access data, the selected safe delivery access data needs to be weighted to improve the accuracy and representativeness of the final result. The core concept of this weighted processing is to perform a weighted calculation on the "safe delivery access data" of different apartment types in the same building, using the apartment type similarity as a weighting factor for the second historical transaction record. The safe delivery access data in all eligible second historical transaction records is weighted and averaged, thereby obtaining safe delivery access data with greater reference value.

[0104] The specific implementation method is: for each second historical transaction record, the corresponding safe delivery access data is multiplied by the apartment type similarity corresponding to the record to obtain a weighted value. Subsequently, all weighted values ​​are accumulated and divided by the sum of all apartment type similarities to obtain the weighted safe delivery access data. This method effectively reflects the impact of the similarity between different apartment types on the judgment of product accessibility, making the system more intelligent and accurate in assessing whether a product is deliverable and installable.

[0105] Suppose user X's delivery address is a 3-bedroom, 2-living room apartment with an area of ​​120 square meters. If no other unit in their unit has purchased the same refrigerator when they purchased it, the system needs to refer to the apartment data of other delivery addresses in the same residential complex and building that have purchased the same refrigerator. Assume the following data exists: Apartment 1: 3 bedrooms, 1 living room (110 square meters), 5 years old, with a historical maximum pass-through width of 85 cm for the refrigerator; Apartment 2: 4 bedrooms, 2 living rooms (140 square meters), 5 years old, with a historical maximum pass-through width of 90 cm for the refrigerator. Calculating apartment similarity: Apartment 1 similarity = 1 - |(120-110) / 120| = 91.7%, Apartment 2 similarity = 1 - |(120-140) / 120| = 83.3%. After weighting, the passing width of the refrigerator is: (85cm×0.917+90cm×0.833) / (0.917+0.833)≈87.67cm.

[0106] If no historical transaction records are found for the same community and building, it means that all users in the building have never purchased the product. At this time, the reference scope will be further expanded to other communities.

[0107] As an optimized implementation, this solution introduces a dynamic range adjustment mechanism to improve the rationality and applicability of data matching. This mechanism dynamically sets the search range based on city density, using a tiered approach. For example, the search range is set to 0.5 kilometers in the core areas (high-density areas) of first-tier cities, 1.5 kilometers in ordinary cities or suburbs, and 3 kilometers in rural or low-density areas. This mechanism dynamically filters out other communities within the search range surrounding the current delivery address.

[0108] As another optimized implementation method, in order to avoid data deviation caused by differences in building types, this solution also introduces a building type matching mechanism to select communities with the same building type as the community where the current delivery address is located (which can be located in the knowledge graph or building data, that is, residential data includes building type), such as elevator buildings or staircase buildings, etc., to ensure that the historical transaction records referenced are closer to the actual physical environment of the current delivery address, thereby improving the accuracy and adaptability of safe delivery pass data, and providing stronger data support for subsequent risk assessment and logistics distribution.

[0109] As another optional implementation, this solution further incorporates a developer matching mechanism when expanding to other residential communities for data reference. For example, based on dynamically adjusting the search scope, other residential communities belonging to the same developer as the current delivery address are prioritized as data reference sources. Because residential communities built by the same developer typically share high similarities in architectural design, apartment layout, and public area structure, this strategy helps improve the applicability and accuracy of historical transaction records in the current scenario.

[0110] By introducing one or more of the above mechanisms, the system can screen out more representative safe delivery access data based on the quality of developers, consistency of architectural style, and distance, even in the absence of historical transaction records of the goods operated by users in the same community and building. After obtaining other communities that meet the conditions, this solution will conduct further screening and analysis based on the data of these other communities. For example, the system will give priority to retrieving data from other communities within 1 km of the current delivery address and from the same developer as the community where the current delivery address is located. If no other community data is found within a 1 km range, the search range will be gradually expanded to a maximum of 3 km to find other community data that meet the conditions.

[0111] After determining other communities, the system will obtain the corresponding target buildings (see the following Figure 5 ), and based on historical transaction records related to the location of other residential communities and target buildings, secure delivery access data is extracted from them (this process is the same as the building-level processing and is not repeated here). This process ensures that even if there are no historical transaction records corresponding to the residential community or building at the current delivery address, valuable reference secure delivery access data can still be obtained from external data sources with similar structure and proximity to the location, supporting subsequent risk assessment and logistics decision-making.

[0112] The method provided in the above embodiment, by constructing a multi-level data matching mechanism, obtains refined residential data such as buildings, units, and apartment types from the knowledge graph of the community, and performs step-by-step matching based on the historical transaction records of the goods, giving priority to obtaining safe delivery and access data for the same community, building, unit, and apartment type. If no such data exists, the matching conditions are gradually relaxed until similar building data is screened across communities. This method can still provide reliable safe delivery and access data in the absence of directly adapted data. This method can provide stable and accurate physical environment data at different data coverage levels, significantly improving the accuracy of physical environment data acquisition and system adaptability in cold start and data sparse scenarios.

[0113] See also Figure 5, shows a flow chart of another optional commodity delivery processing method according to an embodiment of the present invention, comprising the following steps:

[0114] S501: Determine the building information, unit information, and household number information of the other community where the product was purchased based on the delivery address in the historical transaction record;

[0115] S502: Obtaining building data corresponding to the determined building information, and obtaining apartment type data corresponding to the determined building information, unit information, and apartment number information from the knowledge graph of the other community;

[0116] S503: Determine the comparison between the current apartment type data and the acquired apartment type data;

[0117] S504: For the current building data and the building data corresponding to the acquired apartment type data, respectively, calculate the building age difference, building distance, and building structure feature matching degree;

[0118] S505: Filter target buildings based on the apartment type comparison, the building age difference, the building distance, and the matching degree of the building structure characteristics.

[0119] This implementation describes the specific process of determining safe delivery access data at the "community level." After selecting other eligible communities, the system can further identify the specific buildings, units, and households within these communities that have purchased the product based on the user's historical transaction records. Subsequently, combined with the knowledge graph of these other communities, the system can obtain the corresponding household data and the building data of the relevant buildings, providing structured, fine-grained physical environment data to support the determination of safe delivery access data.

[0120] Calculating the comparison between the current apartment data and the acquired apartment data may be performed by calculating apartment similarity, which can be calculated based on the aforementioned apartment similarity formula. Additionally, apartment comparison may be determined by analyzing whether the area difference is less than 15% (this value is for example only and is adjustable), whether the number of rooms is consistent, or by combining apartment similarity, whether the area difference is less than 15%, and whether the number of rooms is consistent. This embodiment preferably considers apartment comparisons based on whether the area difference is less than 15% and whether the number of rooms is consistent.

[0121] For the current building data and the building data corresponding to the acquired apartment data, the building age difference, building distance, and building structure feature matching degree are calculated respectively. For the building age difference, assuming that the building age of the current delivery address is 5 years, and the building age of the acquired building data of other communities is 8 years, then the building age difference is 3 years. The building distance can be determined based on the longitude and latitude of the building or other methods known to those skilled in the art. The matching degree of the building structure features refers to the matching of structural features such as the elevator / staircase location and the unit door orientation. The matching degree here can be calculated using methods known to those skilled in the art. For example, if both are elevator buildings, the matching degree is 1.

[0122] Target buildings can be screened based on the comparison of apartment types, building age differences, building distances, and the matching degree of building structural characteristics. A weighted scoring mechanism or a hierarchical screening mechanism can be used.

[0123] In the first embodiment, a weighted scoring mechanism is used. The weighted summation of the pre-configured weights for the apartment type comparison, building age difference, building distance, and building structural feature matching can be performed to obtain the score of each building, and the building with the highest score is selected as the target building.

[0124] Embodiment 2: Hierarchical screening mechanism. Based on the apartment type comparison, relatively similar apartment type data are screened out, such as apartment types with the same number of rooms and an area difference of less than 15%, and then the building data corresponding to these apartment type data are determined. For these building data, the building age difference, structural feature matching degree, and building distance between them and the current building data are further calculated to screen out building data with a building age difference less than a preset building age difference threshold (such as a completion time difference of less than 5 years) and a structural feature matching degree greater than or equal to a preset matching degree threshold. Among the buildings that meet the above conditions, they are sorted from near to far according to the building distance, and the closest building is preferentially selected as the target building to improve the applicability and accuracy of safe delivery and access data.

[0125] As a preferred embodiment, after calculating the building age difference and building distance, the building age decay factor and distance decay factor can also be calculated. Assuming that the weight of the building age difference decays by 5% with each year of increase, the building age decay factor is: 1-0.05×|building age difference|. Assuming that the building age of the current delivery address is 5 years old, and the building age of the acquired building data in other communities is 8 years old, the calculated building age decay factor is: 1-(8-5) ×5% = 85%. The distance decay factor can be calculated using e^(0.2d), where d is the number of kilometers.

[0126] Taking the weighted scoring mechanism as an example, assuming that based on the apartment type information, unit information, and building information of the current delivery address (located in Community A, the residential data is: elevator building, 3 bedrooms and 2 living rooms, completed in 2023), when no matching historical transaction records are found, the search range is determined to be 1.5km based on the delivery address, and other communities within 1.5km around Community A are searched, and the following are screened out: Apartment type 1 in Building b of Community B: 3 bedrooms and 2 living rooms (95% similarity), elevator building, completed in 2020, 0.8km away; Apartment type 2 in Building c of Community C: 4 bedrooms and 2 living rooms (80% similarity), elevator building, completed in 2022, 1.2km away.

[0127] Assume that the candidate building data is calculated using the following formula for comprehensive scores: Score = 0.4 × Apartment Type Similarity + 0.3 × Building Age Decay Factor + 0.2 × Building Type Matching + 0.1 × Distance Decay Factor. Then, the score for Building b in Community B is: 0.4 × 0.95 + 0.3 × 0.85 + 0.2 × 1 + 0.1 × 0.85 = 0.905; the score for Building c in Community C is: 0.4 × 0.8 + 0.3 × 0.95 + 0.2 × 1 + 0.1 × 0.79 = 0.854. Since 0.905 is greater than 0.854, Building b in Community B is selected as the target building.

[0128] The method provided in the above embodiment, by comprehensively comparing multi-dimensional data such as apartment type, building age, building distance and architectural feature matching, can accurately screen out the most suitable target building from other communities in the absence of historical transaction records in the same community, thereby effectively solving the problem of missing physical environment data in the cold start scenario, ensuring the accuracy and reliability of subsequent product adaptation evaluation, further enhancing user experience and reducing the risk of mispurchase, and having greater practicality and promotion value.

[0129] As an optional implementation, this solution also provides a standardized interface linkage mechanism, which uses a standardized API (Application Programming Interface) interface to achieve millisecond-level data synchronization between the grid system, order center, and customer service system. It supports real-time acquisition of key fields such as physical environment data, product data, historical transaction records, and risk status, ensuring that all links share unified distribution risk assessment results and eliminate information silos.

[0130] As an optional implementation, this solution also incorporates a knowledge graph update mechanism to enhance the system's intelligence and continuous learning capabilities. Upon completing a measurement task, the rider or user will transmit real-world physical environment data. The system then updates the knowledge graph for the corresponding apartment type based on this real-world physical environment data, creating a positive "measure once, reuse multiple times" cycle that continuously enhances the solution's risk prediction capabilities.

[0131] See also Figure 6 , which shows a schematic diagram of the main process of a specific commodity delivery processing method provided by an embodiment of the present invention, including:

[0132] When a user browses a product detail page, the system responds by obtaining the product data and specifications. If the specifications meet the pre-set criteria for large items, the product is determined to be a large item and the user's current delivery address is obtained, which can be the default delivery address or a newly added delivery address by the user before browsing the product. Furthermore, the system can call the home improvement system to obtain the product's dimensions, including external dimensions and packaging dimensions.

[0133] If the system detects that the current delivery address is missing one or more of the community information, building information, unit information, and household number information, a prompt will pop up to inform the user of the missing information and provide a modification option to guide the user to click the modification option to modify the current delivery address. After detecting that there is no missing information in the current delivery address, the system determines the physical environment data based on the current delivery address. Specifically, it can integrate rider measurement data, user-reported data, and safe delivery pass data obtained by reverse deduction based on historical transaction records, and use multi-source data fusion and credibility weighting mechanism to obtain the final physical environment data. It is also possible to consider only data from one or more sources here, and there is no restriction on this.

[0134] After obtaining the physical environment data, the product category is determined and the preset safety margin for that category is obtained. The cumulative value of the product size data and the safety margin is calculated. If this cumulative value is greater than the safe delivery pass data, the delivery risk assessment result is determined to be high risk. If the risk is high, a warning message is displayed, such as a red warning bar, or an AR simulation placement option is provided. In response to the user clicking on this option, an AR image is generated and displayed based on the product data and physical environment data for the user to simulate placement. The user can also be guided to make an appointment for a free measurement service, generate a measurement work order and assign it to a rider, and then receive the actual measured physical environment data uploaded by the rider. If the risk is low, the product information is displayed normally.

[0135] If the delivery risk assessment result is low risk, when the system detects that the user has added an item to the shopping cart, it is necessary to obtain the latest delivery address, which can be the default delivery address or the delivery address newly added by the user before adding the item to the shopping cart. If the latest delivery address is different from the delivery address determined when the user browsed the item, the physical environment data must be re-determined; if they are the same, the physical environment data determined during browsing will be used. The risk assessment model is called to calculate the risk probability value based on the physical environment data and the item data. If the risk probability value is greater than or equal to the preset risk probability threshold, it is judged as high risk and a pop-up window is forced to prompt the risk. Otherwise, it is low risk. In the low-risk case, if the user clicks the order payment option, the order is allowed to be placed normally.

[0136] If the user chooses to pay and place an order, the latest delivery address, i.e., the one selected by the user, must be obtained. If the previous action before paying and placing an order was browsing, the system will determine whether the current latest delivery address is consistent with the one determined during browsing. If the previous action was adding the item to the shopping cart, the system will determine whether the current latest delivery address is consistent with the one determined when adding the item to the shopping cart. If they are inconsistent, the physical environment data must be re-determined, and the delivery risk must be reassessed using the same risk assessment method used for browsing. If the delivery risk assessment result is high, a pop-up window will be displayed to indicate the risk and provide corresponding risk solutions, including but not limited to modifying the product, scheduling a measurement service, and confirming the risk before continuing the order.

[0137] If the user chooses to modify the product, the system automatically recommends compatible products in the same category that match the current physical environment data, based on the product's category. These products are presented in a list for the user to choose from. Once the user selects a product from the recommended list, the system automatically replaces the product while preserving the original product's context. If the user selects the scheduled measurement option, the system automatically generates a measurement order based on the current delivery address and dispatches a delivery driver to collect physical environment data on-site, providing more accurate information for subsequent evaluation and decision support. This physical environment data can be used to update the knowledge graph for the neighborhood where the delivery address is located. If the user chooses to proceed with the order after fully understanding the risks, they will be required to sign an enhanced risk disclosure agreement. After confirming the user's signature, the system will allow the user to proceed with the order and mark the resulting order as high-risk. The order will then be submitted to the customer service system for specialized fulfillment follow-up and service assurance. The system will also generate a composite work order, including the measurement task and delivery schedule, and assign it to the delivery driver. This reduces duplicate visits and improves fulfillment efficiency and user satisfaction.

[0138] This solution achieves several innovative breakthroughs in e-commerce risk assessment. For the first time, it deeply integrates physical environmental data into the core decision-making process of product trading, serving as a key factor in determining product suitability. This overcomes the limitations of traditional systems that rely solely on product attributes and user behavior for risk assessment, significantly improving the objectivity and accuracy of risk identification. Furthermore, by building a multi-level data backup mechanism and integrating a spatial transfer learning algorithm, the system can still obtain effective physical environmental data based on similar apartment or building types, even in cold-start scenarios such as new users or new communities, where environmental data is scarce. Furthermore, this solution embeds real-time risk assessment and intervention mechanisms at key points in the shopping process, achieving a closed-loop control chain from user perception, risk alerts, operational guidance, to fulfillment follow-up, shifting risk control from post-processing to pre-emptive prevention. Finally, by establishing a dual-path decision-making system combining a rules engine and a machine learning model, the rules engine ensures efficient response to high-frequency requests, while the machine learning model continuously optimizes risk prediction accuracy, thus balancing system performance with the needs of intelligent upgrades.

[0139] See also Figure 7 , which shows a schematic diagram of the main modules of a commodity distribution processing device 700 provided by an embodiment of the present invention, including:

[0140] An acquisition module 701 is configured to acquire product data of a product and a current delivery address of the user in response to a transaction-related operation performed by the user on the product;

[0141] An evaluation module 702 is configured to determine physical environment data corresponding to the current delivery address, and determine a delivery risk evaluation result of the product based on the product data and the physical environment data;

[0142] The processing module 703 is used to display the preset risk treatment plans for the transaction-related operations when the delivery risk assessment result is high risk, and execute the selected risk treatment plan in response to the user's selection of one of the risk treatment plans.

[0143] In the embodiment of the present invention, when the transaction-related operation is a browsing operation, the processing module 703 includes the following situations:

[0144] In response to a user selecting an alert message option, displaying the alert message;

[0145] In response to a user selecting an AR simulation placement option, generating AR images based on the product data and the physical environment data, respectively, so that the user performs a simulated placement based on the generated AR images;

[0146] In response to the user's selection of the appointment measurement option, a measurement order is generated based on the current delivery address and assigned to a rider to receive the physical environment data uploaded by the rider.

[0147] In the implementation device of the present invention, when the transaction-related operation is an add-to-cart operation, the processing module 703 is configured to: display a warning message in response to the user selecting the warning message option.

[0148] In the embodiment of the present invention, when the transaction-related operation is an order payment operation, the processing module 703 includes the following situations:

[0149] In response to a user selecting to modify a product option, determining and displaying other products in the same category as the product that are compatible with the physical environment data; in response to a user selecting one of the other products, executing an order and payment process based on the selected other product;

[0150] In response to a user selecting a measurement reservation option, generating a measurement order based on the current delivery address and assigning the order to a rider to receive the physical environment data uploaded by the rider;

[0151] In response to the user's selection of the option to confirm the risk and continue placing the order, the order payment process is executed, and the generated order is marked as a high-risk order and submitted to the customer service system for fulfillment follow-up.

[0152] In the implementation device of the present invention, the physical environment data includes safe delivery pass data, and the evaluation module 702 is used to: determine the category to which the product belongs, and obtain the safety margin preset for the category; obtain size data from the product data, and calculate the cumulative value of the size data and the safety margin; in response to the cumulative value being greater than the safe delivery pass data, determine that the delivery risk assessment result is high risk.

[0153] In the implementation device of the present invention, the evaluation module 702 is used to: extract spatial features from the physical environment data; extract product attributes from the product data of the product; generate cross-features based on the product attributes and spatial features; determine the category to which the product belongs, and obtain the return rate of products in the same category within a preset time period; input the spatial features, the product attributes, the cross-features and the return rate into a preset risk assessment model to obtain a risk probability value; in response to the risk probability value being greater than or equal to a preset risk probability threshold, determine that the distribution risk assessment result is high risk.

[0154] In the embodiment of the present invention, the evaluation module 702 includes a determination module configured to:

[0155] Determining a previous transaction-related operation performed by the user on the product before the transaction-related operation, so as to determine a previous delivery address obtained when the user performed the previous transaction-related operation on the product;

[0156] In response to the current delivery address being the same as the previous delivery address, using the physical environment data determined based on the previous delivery address as the physical environment data corresponding to the current delivery address;

[0157] In response to the absence of the previous transaction-related operation or the current delivery address being different from the previous delivery address, physical environment data corresponding to the current delivery address is determined.

[0158] In the implementation device of the present invention, the determination module includes one or more of the following situations: generating a measurement order based on the current delivery address and assigning it to a rider to receive the physical environment data collected from the current delivery address uploaded by the rider; receiving the physical environment data uploaded by the user; and determining the physical environment data corresponding to the current delivery address based on the historical transaction records of the product.

[0159] In the implementation device of the present invention, the current delivery address includes community information, building information, unit information, and household number information; the determination module is used to:

[0160] Obtaining residential data corresponding to the current delivery address from a knowledge graph corresponding to the community information; wherein the residential data includes current building data corresponding to the building information, and current apartment type data corresponding to the building information, unit information, and apartment number information;

[0161] In order of household number, unit, building, and community granularity from fine to coarse, based on the historical transaction records of the goods, secure delivery access data corresponding to the residential data is obtained.

[0162] In the implementation device of the present invention, the determination module is used to obtain the safe delivery access data corresponding to the residential data based on the historical transaction records of the commodity at a cell granularity, and the process includes:

[0163] Based on the historical transaction records of the product, determine other community information, and filter out the target building from the other communities according to the current apartment type data and the current building data;

[0164] Obtaining, from the historical transaction records of the product, a first historical transaction record corresponding to other communities and target buildings, obtaining other delivery addresses from the first historical transaction record, and obtaining other apartment type data for the other delivery addresses;

[0165] Calculate the apartment type similarity between the current apartment type data and other apartment type data, filter out a second historical transaction record whose apartment type similarity is greater than or equal to a preset similarity threshold from the first historical transaction record, and use the safe delivery pass data in the second historical transaction record as the safe delivery pass data for the current delivery address.

[0166] In the implementation device of the present invention, the determining module is used to:

[0167] Determine the building information, unit information, and household number information of the other community where the product was purchased based on the delivery address in the historical transaction record;

[0168] Obtaining, from the knowledge graph of the other community, building data corresponding to the determined building information, and obtaining apartment type data corresponding to the determined building information, unit information, and apartment number information;

[0169] Determine the comparison between the current apartment data and the acquired apartment data;

[0170] For the current building data and the building data corresponding to the acquired apartment data, calculate the building age difference, building distance and building structure feature matching degree respectively;

[0171] Target buildings are screened based on the apartment type comparison, the building age difference, the building distance, and the matching degree of the building structural features.

[0172] In addition, the specific implementation content of the device in the embodiment of the present invention has been described in detail in the above method, so the repeated content will not be described again here.

[0173] Figure 8 An exemplary system architecture 800 to which embodiments of the present invention may be applied is shown, including terminal devices 801 , 802 , 803 , a network 804 and a server 805 (only an example).

[0174] Terminal devices 801, 802, and 803 can be various electronic devices with display screens and support web browsing, and are installed with various communication client applications. Users can use terminal devices 801, 802, and 803 to interact with server 805 through network 804 to receive or send messages, etc.

[0175] The network 804 is used to provide a medium for communication links between the terminal devices 801, 802, 803 and the server 805. The network 804 may include various connection types, such as wired or wireless communication links or fiber optic cables.

[0176] Server 805 can be a server that provides various services, such as a backend management server that supports shopping websites browsed by users using terminal devices 801, 802, and 803 (for example only). The backend management server can analyze and process received data such as product information query requests, and feed back the processing results (such as target push information and product information—for example only) to the terminal device. It should be noted that the methods provided in the embodiments of the present invention are generally executed by server 805, and accordingly, the device is generally located in server 805.

[0177] It should be understood that Figure 8 The number of terminal devices, networks and servers in the embodiment is merely illustrative. Any number of terminal devices, networks and servers may be provided as required.

[0178] Reference below Figure 9 , which shows a schematic structural diagram of a computer system 900 of a terminal device suitable for implementing an embodiment of the present invention. Figure 9 The terminal device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0179] like Figure 9 As shown, computer system 900 includes a central processing unit (CPU) 901, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 902 or a program loaded from a storage unit 908 into a random access memory (RAM) 903. Various programs and data required for the operation of system 900 are also stored in RAM 903. CPU 901, ROM 902, and RAM 903 are connected to each other via a bus 904. An input / output (I / O) interface 905 is also connected to bus 904.

[0180] The following components are connected to the I / O interface 905: an input section 906 including a keyboard, a mouse, and the like; an output section 907 including devices such as a cathode ray tube (CRT), a liquid crystal display (LCD), and speakers; a storage section 908 including devices such as a hard disk; and a communication section 909 including a network interface card such as a LAN card or a modem. The communication section 909 performs communication processing via a network such as the Internet. A drive 910 is also connected to the I / O interface 905 as needed. Removable media 911, such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, is installed in the drive 910 as needed, so that computer programs read therefrom can be installed in the storage section 908 as needed.

[0181] In particular, according to embodiments disclosed herein, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed herein include a computer program product comprising a computer program embodied on a computer-readable medium, the computer program containing program code for executing the methods illustrated in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 909 and / or installed from removable media 911. When executed by central processing unit (CPU) 901, the computer program performs the aforementioned functions defined in the system of the present invention.

[0182] It should be noted that the computer-readable medium described in the present invention may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. Computer-readable storage media may include, for example, but not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or components, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection having one or more conductors, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device, or component. In the present invention, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such a propagated data signal may take a variety of forms, including, but not limited to, electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device. Program code embodied on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wireline, optical fiber cable, RF, or any suitable combination thereof.

[0183] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or a part of code, and the above-mentioned module, program segment, or a part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0184] The modules described in the embodiments of the present invention may be implemented in software or hardware. The modules described may also be provided in a processor. For example, they may be described as comprising an acquisition module, an evaluation module, and a processing module. The names of these modules do not, in some cases, limit the modules themselves. For example, the evaluation module may also be described as a "risk control module."

[0185] As another aspect, the present invention further provides a computer-readable medium, which may be included in the device described in the above embodiments, or may exist independently and not incorporated into the device. The computer-readable medium carries one or more programs, which, when executed by the device, cause the device to perform any of the above-described methods for processing product delivery.

[0186] The computer program product of the present invention includes a computer program, which, when executed by a processor, implements the commodity delivery processing method in the embodiment of the present invention.

[0187] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may occur depending on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A commodity distribution processing method, characterized in that: include: In response to a transaction-related operation performed by a user on a product, obtaining product data of the product and obtaining the user's current delivery address; Determining physical environment data corresponding to the current delivery address, and determining a delivery risk assessment result of the product based on the product data and the physical environment data; In the case where the delivery risk assessment result is high risk, the preset risk handling solutions for the transaction-related operations are displayed, and in response to the user's selection operation of one of the risk handling solutions, the selected risk handling solution is executed.

2. The method according to claim 1, characterized in that In the case where the transaction-related operation is a browsing operation, the executing of the selected risk treatment solution in response to the user selecting one of the risk treatment solutions includes one of the following situations: In response to a user selecting an alert message option, displaying the alert message; In response to a user selecting an AR simulation placement option, generating AR images based on the product data and the physical environment data, respectively, so that the user performs a simulated placement based on the generated AR images; In response to the user's selection of the appointment measurement option, a measurement order is generated based on the current delivery address and assigned to a rider to receive the physical environment data uploaded by the rider.

3. The method according to claim 1, characterized in that In a case where the transaction-related operation is an add-to-cart operation, executing the selected risk handling solution in response to a user's selection operation of one of the risk handling solutions includes: displaying warning information in response to a user's selection operation of a warning information option.

4. The method according to claim 1, wherein In the case where the transaction-related operation is an order payment operation, the executing of the selected risk treatment solution in response to the user selecting one of the risk treatment solutions includes one of the following situations: In response to a user selecting to modify a product option, determining and displaying other products in the same category as the product that are compatible with the physical environment data; in response to a user selecting one of the other products, executing an order and payment process based on the selected other product; In response to a user selecting a measurement reservation option, generating a measurement order based on the current delivery address and assigning the order to a rider to receive the physical environment data uploaded by the rider; In response to the user's selection of the option to confirm the risk and continue placing the order, the order payment process is executed, and the generated order is marked as a high-risk order and submitted to the customer service system for fulfillment follow-up.

5. The method according to claim 1, wherein The physical environment data includes safe delivery passage data, and determining the delivery risk assessment result of the commodity based on the commodity data and the physical environment data includes: Determine the category to which the product belongs, and obtain a preset safety margin for the category; Acquire dimension data from the commodity data, and calculate a cumulative value of the dimension data and the safety margin; In response to the accumulated value being greater than the safe delivery pass data, the delivery risk assessment result is determined to be high risk.

6. The method according to claim 1, wherein Determining a distribution risk assessment result of the commodity based on the commodity data and the physical environment data includes: extracting spatial features from the physical environment data; extracting product attributes from product data of the product; generating cross features based on the commodity attributes and spatial features; Determine the category to which the product belongs, and obtain the return rate of products in the same category within a preset period of time; Inputting the spatial features, the commodity attributes, the cross features, and the return rate into a preset risk assessment model to obtain a risk probability value; In response to the risk probability value being greater than or equal to a preset risk probability threshold, the delivery risk assessment result is determined to be high risk.

7. The method according to claim 1, characterized in that The determining of the physical environment data corresponding to the current delivery address includes: Determining a previous transaction-related operation performed by the user on the product before the transaction-related operation, so as to determine a previous delivery address obtained when the user performed the previous transaction-related operation on the product; In response to the current delivery address being the same as the previous delivery address, using the physical environment data determined based on the previous delivery address as the physical environment data corresponding to the current delivery address; In response to the absence of the previous transaction-related operation or the current delivery address being different from the previous delivery address, physical environment data corresponding to the current delivery address is determined.

8. The method according to claim 1 or 7, characterized in that The determining of the physical environment data corresponding to the current delivery address includes one or more of the following situations: Generate a measurement order based on the current delivery address and assign it to the rider, so as to receive the physical environment data collected from the current delivery address uploaded by the rider; Receive physical environment data uploaded by users; Based on the historical transaction records of the product, physical environment data corresponding to the current delivery address is determined.

9. The method according to claim 8, characterized in that The current delivery address includes community information, building information, unit information, and household number information; The determining of the physical environment data corresponding to the current delivery address based on the historical transaction records of the product includes: Obtaining residential data corresponding to the current delivery address from a knowledge graph corresponding to the community information; wherein the residential data includes current building data corresponding to the building information, and current apartment type data corresponding to the building information, unit information, and apartment number information; In order of household number, unit, building, and community granularity from fine to coarse, based on the historical transaction records of the goods, secure delivery access data corresponding to the residential data is obtained.

10. The method according to claim 9, characterized in that At the cell level, based on the historical transaction records of the product, secure delivery access data corresponding to the residential data is obtained, including: Based on the historical transaction records of the product, determine other community information, and filter out the target building from the other communities according to the current apartment type data and the current building data; Obtaining, from the historical transaction records of the product, a first historical transaction record corresponding to other communities and target buildings, obtaining other delivery addresses from the first historical transaction record, and obtaining other apartment type data for the other delivery addresses; Calculate the apartment type similarity between the current apartment type data and other apartment type data, filter out a second historical transaction record whose apartment type similarity is greater than or equal to a preset similarity threshold from the first historical transaction record, and use the safe delivery pass data in the second historical transaction record as the safe delivery pass data for the current delivery address.

11. The method according to claim 10, characterized in that The step of selecting a target building from the other communities based on the current apartment type data and the current building data includes: Determine the building information, unit information, and household number information of the other community where the product was purchased based on the delivery address in the historical transaction record; Obtaining, from the knowledge graph of the other community, building data corresponding to the determined building information, and obtaining apartment type data corresponding to the determined building information, unit information, and apartment number information; Determine the comparison between the current apartment data and the acquired apartment data; For the current building data and the building data corresponding to the acquired apartment data, calculate the building age difference, building distance and building structure feature matching degree respectively; Target buildings are screened based on the apartment type comparison, the building age difference, the building distance, and the matching degree of the building structural features.

12. A commodity distribution processing device, characterized in that: include: an acquisition module, configured to acquire product data of a product and a current delivery address of the user in response to a transaction-related operation performed by the user on the product; an assessment module, configured to determine physical environment data corresponding to the current delivery address, and determine a delivery risk assessment result of the product based on the product data and the physical environment data; The processing module is used to display the preset risk processing solutions for the transaction-related operations when the delivery risk assessment result is high risk, and execute the selected risk processing solution in response to the user's selection operation of one of the risk processing solutions.

13. An electronic device, characterized in that: include: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 11.

14. A computer-readable medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 11 is implemented.

15. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 11 is implemented.

Images