User behavior-based merchandise display method, device, equipment and medium

Abstract

The application provides a commodity display method and device based on user behavior, equipment and medium, belongs to the technical field of vending machines, the method is applied to a vending machine with multiple channels, comprising: obtaining a user transaction data summary table and a user behavior data summary table with a timestamp in a statistical period, and the goods and channels are allocated three-dimensional coordinate codes; fusion and preprocessing the two tables to obtain a fusion data table, determining the support, confidence and purchase conversion rate of each type of commodity, combining the association rule to generate an association rule report, and then generating a commodity display adjustment scheme; wherein the purchase conversion rate is based on the attention index, and the attention index is calculated based on the effective stay time and the number of visual focus times. The commodity display method, device, equipment and medium based on user behavior provided by the application can realize accurate matching of commodity display, user behavior and purchase demand.

Description

Technical Field

[0001] This application belongs to the field of vending machine technology, and more specifically, relates to a method, device, equipment and medium for displaying goods based on user behavior. Background Technology

[0002] Multi-channel vending machines used in commercial districts and transportation hubs need to adapt their product displays to user consumption needs to improve operational efficiency. Current product display optimization for vending machines often relies solely on single user transaction data, failing to incorporate actual user behavior data during the product selection process or effectively quantify the correlation between product attention and purchase. This results in a lack of comprehensive and rigorous data analysis support for display adjustments, relying primarily on manual experience. This approach fails to accurately reflect users' actual browsing and purchasing habits, and the formulation and adjustment of display layouts lack a scientific basis. Ultimately, this leads to vending machine product displays that do not align with actual user needs, failing to achieve a precise match between product display and user behavior and purchasing demands. Summary of the Invention

[0003] The purpose of this application is to provide a user behavior-based product display method, apparatus, equipment, and medium capable of accurately matching product display with user behavior and purchasing needs. To achieve the above objective, the technical solution provided by this application is as follows: Firstly, a user behavior-based product display method is provided for vending machines, which include multiple product aisles. The method includes: Obtain the first and second data tables of the vending machine. The first data table is a summary table of user transaction data for products within the statistical period, and the second data table is a summary table of user behavior data generated when users select products within the statistical period. Each product and multiple product channels in the vending machine are assigned a three-dimensional coordinate code. The data in the first and second data tables carries a timestamp. The data in the first data table and the second data table are merged, and the merged data is preprocessed to obtain a merged data table; Based on the integrated data table, the support, confidence and purchase conversion rate of each type of product are determined; Based on the support, confidence, and purchase conversion rate of various products, and combined with the association rules between products and channels, a report on association rules between products and channels is generated. The association rule is a combination of three-dimensional coordinate codes of products and channels that simultaneously meet the predefined support threshold, confidence threshold, and purchase conversion rate threshold. Generate product display adjustment plans based on association rule reports; For each product, The support level for each type of product is calculated based on the number of times the product is purchased in a specific product channel and the total number of times the product is purchased across all product channels. The confidence level of each type of product is calculated based on the number of times the product is purchased in a certain sales channel and the total number of times all products in that sales channel are purchased. The purchase conversion rate for each type of product is calculated based on the number of times the product is purchased and the product's attention index. The second data table includes at least the effective dwell time and the number of times the product's eyes are focused. The attention index is calculated based on the effective dwell time and the number of times the product's eyes are focused.

[0004] Secondly, a user behavior-based product display device is provided for use in vending machines, which include multiple product aisles. The method includes: The data acquisition module is used to acquire the first data table and the second data table of the vending machine. The first data table is a summary table of user transaction data for products within the statistical period, and the second data table is a summary table of user behavior data generated when users select products within the statistical period. Each product and multiple product channels in the vending machine are assigned a three-dimensional coordinate code. The data in the first and second data tables carries a timestamp. The data fusion module is used to merge the data in the first data table and the second data table, and to preprocess the merged data to obtain a merged data table. The feature extraction module is used to determine the support, confidence, and purchase conversion rate of various products based on the fused data table; The report generation module is used to generate a report on the association rules between products and channels based on the support, confidence, and purchase conversion rate of various products, combined with the association rules between products and channels. The association rules are combinations of three-dimensional coordinate codes of products and channels that simultaneously meet the predefined support threshold, confidence threshold, and purchase conversion rate threshold. The product adjustment module is used to generate product display adjustment plans based on association rule reports; For each product, The support level for each type of product is calculated based on the number of times the product is purchased in a specific product channel and the total number of times the product is purchased across all product channels. The confidence level of each type of product is calculated based on the number of times the product is purchased in a certain sales channel and the total number of times all products in that sales channel are purchased. The purchase conversion rate for each type of product is calculated based on the number of times the product is purchased and the product's attention index. The second data table includes at least the effective dwell time and the number of times the product's eyes are focused. The attention index is calculated based on the effective dwell time and the number of times the product's eyes are focused.

[0005] Thirdly, embodiments of this application also provide an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the user behavior-based merchandise display method provided by any possible implementation of the first aspect.

[0006] Fourthly, embodiments of this application also provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the user behavior-based product display method provided by any possible implementation of the first aspect.

[0007] The beneficial effects of the technical solution provided in this application are as follows: The product display method, apparatus, equipment, and medium based on user behavior provided in this application, compared with related technologies, achieves precise temporal and spatial correlation between transaction data and behavioral data by acquiring time-stamped user transaction and behavior dual data tables and combining them with product and channel-specific three-dimensional coordinate codes. This solves the one-sidedness of analyzing display problems from a single data dimension. The dual data tables are merged and preprocessed, invalid data is removed, and valid information is integrated, ensuring the accuracy of support, confidence, and purchase conversion rate calculations. The purchase conversion rate is derived from the attention index calculated based on effective dwell time and number of eye-focusing times, accurately reflecting the conversion relationship between user attention and purchase. By combining three indicators with preset thresholds to filter effective product-channel correlation rules and generate a correlation rule report, the generation of display adjustment plans has a rigorous data basis, abandoning the blindness of experience-based displays. Display plans are generated based on this report, providing clear execution guidelines for product display adjustments. Overall, this achieves precise matching between product display and user behavior and purchasing needs. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments of this application will be briefly introduced below.

[0009] Figure 1 A flowchart illustrating the product display method based on user behavior provided in this application embodiment; Figure 2 A structural block diagram of a merchandise display device based on user behavior provided in an embodiment of this application; Figure 3 A schematic block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0010] The embodiments of this application are described below with reference to the accompanying drawings. It should be understood that the embodiments described below with reference to the accompanying drawings are exemplary descriptions for explaining the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions of the embodiments of this application.

[0011] Those skilled in the art will understand that, unless otherwise stated, the singular forms “a,” “an,” and “the” used herein may also include the plural forms. It should be further understood that the terms “comprising” and “including” as used in embodiments of this application mean that the corresponding feature can be implemented as the presented feature, information, data, step, operation, element, and / or component, but do not exclude implementation as other features, information, data, step, operation, element, component, and / or combinations thereof supported by the art. It should be understood that when we say that an element is “connected” or “coupled” to another element, the one element can be directly connected or coupled to the other element, or it can mean that the one element and the other element establish a connection relationship through an intermediate element. Furthermore, “connected” or “coupled” as used herein can include wireless connection or wireless coupling. The term “and / or” as used herein indicates at least one of the items defined by the term; for example, “A and / or B” can be implemented as “A,” or as “B,” or as “A and B.” When describing multiple (two or more) items, if the relationship between the multiple items is not explicitly defined, the multiple items can refer to one, several or all of the multiple items. For example, the description of "parameter A includes A1, A2, A3" can be implemented as parameter A includes A1 or A2 or A3, or it can be implemented as parameter A includes at least two of the three items A1, A2 and A3.

[0012] It is understood that in the embodiments of this application, data such as user information are involved. When the embodiments of this application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of related data must comply with relevant laws, regulations and standards.

[0013] For example, when this application embodiment is applied to a large vending machine in a shopping district, the user's explicit permission will be obtained first through a pop-up window on the vending machine's operation panel. Video behavior data of the product aisle area will only be collected after the user agrees. Personal identification information such as faces will be removed immediately after collection. The data will only be used for product display optimization, and the user can withdraw permission at any time on the operation panel. The entire process of data collection and processing complies with relevant data security laws and regulations.

[0014] This application provides a product display method based on user behavior, applied to a vending machine. The vending machine includes multiple product channels, and this method can be executed by an electronic device, such as... Figure 1 As shown, the method may include: S101: Obtain the first data table and the second data table of the vending machine. The first data table is a summary table of user transaction data for goods within the statistical period, and the second data table is a summary table of user behavior data generated when users select goods within the statistical period. Each product and multiple product channels in the vending machine are assigned a three-dimensional coordinate code. The data in the first data table and the second data table carry timestamps.

[0015] In this embodiment, the vending machine serves as the application carrier, equipped with multiple product aisles to display various goods. It collects user transaction data and purchasing behavior data to enable real-time collection and uploading of product display, sales, and user behavior data. The product aisles in this vending machine employ a layered, column-based, and row-based three-dimensional layout, facilitating precise positioning and product display adjustments. The three-dimensional coordinate encoding uses an XYZ encoding format. The X-axis represents the number of left and right aisle columns, ranging from 1 to N, where N is the total number of horizontal aisles in the vending machine. The Y-axis represents the number of vertical aisle layers, ranging from 1 to M, where M is the total number of vertical aisles in the vending machine. The Z-axis represents the number of front and rear aisle rows, with 1 for the front row and 2 for the rear row. Each product and each aisle is assigned a unique three-dimensional coordinate code. After encoding, the codes are stored in the vending machine's local database and synchronized to the cloud-based data analysis system, ensuring accurate association between products and aisles.

[0016] The statistical period is a pre-set fixed time length, which serves as a unified time benchmark for the collection, aggregation, and calculation of user transaction data and user behavior data, ensuring the relevance and comparability of the data; it can be set to 7 days by default.

[0017] A timestamp is a time identifier attached to each data entry in the first and second data tables. It is used to record the specific time when each data entry was generated. The timestamp can be a millisecond-level timestamp, and the format can be YYYY-MM-DD-HH-MM-SS-SSS. It is used to ensure that the time base of the data in the first and second data tables is consistent, which facilitates subsequent data fusion and analysis. The timestamp can be generated by the main control module of the vending machine.

[0018] The first data table is a summary table of user transaction data for goods in the vending machine during the statistical period. Its core purpose is to record relevant information about users' purchases of goods, and it is the core transaction data carrier for analyzing product purchases and calculating relevant indicators. The second data table is a summary table of user behavior data generated when users select goods in front of the vending machine during the statistical period. Its core purpose is to record relevant behavioral information of users during the selection process, including at least the effective dwell time and the number of times the eyes are focused. It is the core behavioral data carrier for analyzing users' attention to products and calculating the attention index.

[0019] In this embodiment, the vending machine's first and second data tables are first obtained. The first data table is a summary table of user transaction data for goods within the statistical period, recording relevant transaction information of user purchases. The second data table is a summary table of user behavior data generated when users select goods within the statistical period, recording relevant user browsing and attention behavior information. To achieve precise association between goods and aisles, each goods and each aisle of the vending machine is assigned a unique three-dimensional coordinate code. Simultaneously, all data in both the first and second data tables carries a timestamp to ensure consistency in the time base of the two types of data, avoiding distortion in subsequent analysis due to time discrepancies. The aim is to collect all the raw data required for display optimization, providing fundamental support for all subsequent analysis stages.

[0020] S102: Merge the data in the first data table and the second data table, and preprocess the merged data to obtain a merged data table.

[0021] In this embodiment, the fused data table is a comprehensive data table obtained by merging and preprocessing the first and second data tables. It integrates user transaction data and user behavior data, eliminating invalid data and differences in data dimensions.

[0022] In this embodiment, after obtaining two types of original data tables, the user transaction data in the first data table and the user behavior data in the second data table are correlated with the product as the core association dimension, thus achieving the fusion of the two types of data. Simultaneously, the fused original data undergoes preprocessing, primarily including removing abnormal data such as test transactions and invalid behaviors, standardizing the data format, and eliminating differences in the data units of different indicators, ultimately resulting in a fused data table. Its purpose is to integrate scattered data resources, improve data quality, and form a comprehensive data carrier that can simultaneously reflect user transaction behavior and browsing / attention behavior, laying the foundation for subsequent core indicator calculations and association rule mining.

[0023] S103: Determine the support, confidence and purchase conversion rate for each type of product based on the fused data table.

[0024] In this embodiment, support is a metric used for each product category to measure the purchase popularity of that product in a specific sales channel. Confidence is a metric used for each product category to measure the suitability of that product for a specific sales channel. Purchase conversion rate is a metric used for each product category to measure the probability that a user's interest in a product translates into an actual purchase.

[0025] In this embodiment, based on the preprocessed fused data table, three indicators—support, confidence, and purchase conversion rate—are calculated for each product category. These three indicators quantify the matching degree between products and sales channels, and the correlation between user attention and purchase from different dimensions. Support reflects the purchase popularity of a product in a specific sales channel, confidence reflects the suitability of a specific sales channel for the product, and purchase conversion rate reflects the ability of user attention to convert into purchase. The three indicators work together to form the quantitative core of association rule mining, ensuring that subsequent rule mining has clear data support.

[0026] S104: Based on the support, confidence, and purchase conversion rate of various products, and combined with the association rules between products and channels, generate an association rule report between products and channels; the association rule is a combination of three-dimensional coordinate codes of products and channels that simultaneously meet the predefined support threshold, confidence threshold, and purchase conversion rate threshold. For each product, The support level for each type of product is calculated based on the number of times the product is purchased in a specific product channel and the total number of times the product is purchased across all product channels. The confidence level of each type of product is calculated based on the number of times the product is purchased in a certain sales channel and the total number of times all products in that sales channel are purchased. The purchase conversion rate for each type of product is calculated based on the number of times the product is purchased and the product's attention index. The second data table includes at least the effective dwell time and the number of times the product's eyes are focused. The attention index is calculated based on the effective dwell time and the number of times the product's eyes are focused.

[0027] In this embodiment, the association rule refers to the effective matching relationship between products and product channels. Specifically, it refers to the combination of three-dimensional coordinate codes of products and product channels that simultaneously meet predefined support thresholds, confidence thresholds, and purchase conversion rate thresholds. This is the core basis for display adjustments. The predefined support threshold is a pre-set support standard used to determine whether a product and product channel combination constitutes a valid association rule. Only when the support of a product and product channel combination reaches this threshold can it be further determined as a valid association. The predefined confidence threshold is a pre-set confidence standard that, in conjunction with the support threshold and purchase conversion rate threshold, is used to filter valid association rules between products and product channels. The predefined purchase conversion rate threshold is a pre-set purchase conversion rate standard and is one of the thresholds for filtering valid association rules between products and product channels, ensuring that the association rule can reflect the conversion ability of user attention to purchase.

[0028] The association rule report is generated based on the integrated data table, three indicators, and association rules. It is used to present effective product-shopping channel association combinations, product-shopping channel matching characteristics, and other content, providing a direct basis for generating product display adjustment plans.

[0029] In this embodiment, the effective dwell time refers to the cumulative time when a user's gaze is continuously focused on a certain product's corresponding aisle for ≥1 second, excluding invalid data with a dwell time <1 second; the number of gaze focuses refers to the number of times a user's gaze effectively dwells on a certain product's corresponding aisle within the statistical period, with each effective dwell time counted as 1 time; the attention index is a parameter that quantifies the user's attention to the product, and the calculation formula is Attention Index = Effective Dwell Time × Number of Gaze Focuses. The higher the attention index value, the higher the overall attention to the product.

[0030] In this embodiment, the support score is calculated based on the number of times the product is purchased in a specific aisle and the total number of times the product is purchased in all aisles of the vending machine; the confidence score is calculated based on the number of times the product is purchased in a specific aisle and the total number of times all products in that aisle are purchased; and the purchase conversion rate is calculated based on the number of times the product is purchased and the attention index of the product, wherein the attention index is calculated from the effective dwell time and the number of times the eyes are focused in the second data table. This refined calculation logic ensures the quantitative accuracy and operability of the three indicators, providing a guarantee for the accurate mining of association rules.

[0031] Based on the calculated support, confidence, and purchase conversion rates of various products, and combined with the association rules between products and aisles (i.e., the three-dimensional coordinate code combinations of products and aisles that simultaneously meet predefined support thresholds, confidence thresholds, and purchase conversion rate thresholds), all combinations of products and aisles are screened and organized to extract effective association combinations, ultimately generating a report on the association rules between products and aisles. This report clearly presents which products are suitable for display in which aisles, providing direct theoretical and data support for the generation of subsequent display adjustment plans.

[0032] S105: Generate a product display adjustment plan based on the association rule report.

[0033] In this embodiment, the product display adjustment plan is an actionable plan based on the association rule report, which clarifies the details and basis for adjusting the display position of each product, and is used to guide the actual optimization of the product display in the vending machine.

[0034] In this embodiment, based on the association rule report and combined with the actual aisle layout constraints of the vending machine (such as aisle load-bearing capacity, product category display specifications, etc.), a specific and implementable product display adjustment plan is formulated. This plan clarifies the original display location of each product (corresponding to the original aisle's 3D coordinate code), the target display location (corresponding to the target aisle's 3D coordinate code), and the basis for adjustment (corresponding to the effective association combinations in the association rule report). The aim is to transform the theoretical data in the association rule report into a display optimization plan that can directly guide on-site operations, achieving a precise match between product display and user behavior and purchasing needs.

[0035] As can be seen from the above, this embodiment achieves precise temporal and spatial correlation between transaction data and behavioral data by acquiring time-stamped user transaction and behavior data tables, combined with the unique three-dimensional coordinate codes of products and distribution channels. This solves the one-sidedness of analyzing display issues from a single data dimension. The dual data tables are merged and preprocessed, eliminating invalid data and integrating valid information, ensuring the accuracy of support, confidence, and purchase conversion rate calculations. The purchase conversion rate, based on the attention index calculated from effective dwell time and number of eye-focusing times, accurately reflects the conversion relationship between user attention and purchase. By combining three indicators with preset thresholds to filter effective product-distribution channel correlation rules and generate a correlation rule report, the generation of display adjustment plans has a rigorous data basis, abandoning the blindness of experience-based displays. Display plans are generated based on this report, providing clear execution guidelines for product display adjustments. Overall, this achieves precise matching between product display and user behavior and purchasing needs, improving the scientific and targeted nature of vending machine product display adjustments and laying a data foundation for improving product display optimization effects.

[0036] In one embodiment of this application, a product display adjustment plan is generated based on an association rule report, including: Based on the association rule report, a set of high-frequency purchased products and a set of high-attention, low-purchase products are identified. The set of high-frequency purchased products consists of products with a purchase frequency higher than the first threshold, while the set of high-attention, low-purchase products consists of products with an attention index higher than the second threshold and a purchase conversion rate lower than the third threshold. Based on the set of frequently purchased goods, the set of high-attention but low-purchase goods, and the preset channel adjustment constraints, the three-dimensional coordinate code of the target channel for each goods is determined. Based on the original and target 3D coordinate codes of the products, a product display adjustment plan is generated. The product display adjustment plan includes the product code, the original 3D coordinate code, the target 3D coordinate code, and the basis for adjustment for each product. The constraints for adjusting the cargo lane include: For products belonging to the same category, the corresponding target channel three-dimensional coordinate codes are arranged adjacently or continuously in space.

[0037] In this embodiment, identifying the high-frequency purchase product set and the high-attention, low-purchase product set involves extracting purchase frequency, attention index, and purchase conversion rate data for all products from the association rule report, and comparing them with preset thresholds to complete product classification. The first threshold is a normalized purchase frequency of 0.8, meaning products with a purchase frequency ≥ 0.8 after Min-Max standardization are classified into the high-frequency purchase product set. The second threshold is a normalized attention index of 0.7, and the third threshold is a purchase conversion rate of 0.03, meaning products with an attention index ≥ 0.7 after Min-Max standardization and a purchase conversion rate ≤ 0.03 are classified into the high-attention, low-purchase product set. The channel adjustment constraint is that the three-dimensional coordinate codes of the target channels for products of the same category are spatially adjacent or contiguous. The product categories can include beverages, snacks, and daily necessities. Adjacent or continuous arrangement means that the X, Y, or Z axis coordinate values ​​of the aisle 3D coordinate codes are continuous. For example, X2-Y1-Z1 and X3-Y1-Z1 are adjacent coordinates, and X2-Y1-Z1, X3-Y1-Z1, and X4-Y1-Z1 are continuous coordinates. Determining the target aisle 3D coordinate code involves matching a suitable aisle 3D coordinate code for each product based on the product classification results and aisle adjustment constraints, replacing its original aisle 3D coordinate code. The specific content of the product display adjustment plan includes the product code, the original aisle 3D coordinate code, the target aisle 3D coordinate code, and the adjustment basis. The adjustment basis is the set type to which the product belongs and the aisle adjustment constraints.

[0038] In this embodiment, during the process of generating a product display adjustment plan based on the association rule report, product data is first extracted from the association rule report according to a preset numerical threshold. The products in the vending machine are then accurately classified by feature, dividing them into a set of high-frequency purchased products and a set of high-attention, low-purchase products. The core product objects for display optimization are then identified. Next, the adjustment constraints of the target aisle 3D coordinate codes of the same category of products are combined with the spatially adjacent or continuous aisle adjustment conditions. A suitable target aisle 3D coordinate code is determined for each product. Finally, based on the difference between the original aisle 3D coordinate code and the target aisle 3D coordinate code of each product, a product display adjustment plan containing complete information such as product code, coordinate code, and adjustment basis is generated.

[0039] As can be seen from the above, this embodiment achieves accurate feature classification of goods in the vending machine by setting a preset threshold, enabling the display adjustment to formulate differentiated optimization strategies for goods with different characteristics, thereby improving the pertinence and effectiveness of the display adjustment; the constraint of adjacent or continuous display of goods of the same category avoids browsing confusion during the user's shopping process and improves the user's shopping experience; at the same time, the product display adjustment plan contains complete adjustment information, clearly defining the adjustment basis for each product, which greatly improves the implementation and feasibility of the plan.

[0040] In one embodiment of this application, based on a set of frequently purchased goods, a set of goods with high attention but low purchase, and preset channel adjustment constraints, the target channel three-dimensional coordinate code for each goods is determined, including: From the set of frequently purchased products, select the N products with the highest number of purchases to form the core optimized product set; From the set of high-attention, low-purchase products, select the top M products by attention index to form the core adjustment product set; Based on the preset golden display area mapping table, each product in the core optimized product set is assigned the three-dimensional coordinate code of the aisle corresponding to the golden display area as its target aisle three-dimensional coordinate code. The golden display area is the physical aisle area where the eye focus frequency is higher than the preset frequency threshold, which is determined by statistical analysis of historical user behavior data. For each item in the core product set, assign its target aisle three-dimensional coordinate code as the spatially adjacent aisle three-dimensional coordinate code of the aisle to the prime display area. For other goods not included in the core optimized goods set and the core adjusted goods set, a target three-dimensional coordinate code is assigned to them based on the lane adjustment constraints and the three-dimensional coordinate code of the remaining lanes.

[0041] In this embodiment, N and M are both preset selection quantities. N is 30, meaning that the top 30% of the most frequently purchased items in the high-frequency purchase item set are selected to form the core optimized item set. The value of M is determined based on the actual number of adjacent aisles in the prime display area, ensuring that the core adjusted items can be matched with their corresponding adjacent aisles. The prime display area mapping table is a table pre-generated based on historical user behavior data, containing the three-dimensional coordinate codes of the aisles in the prime display area and the corresponding user gaze focus frequency data. The preset frequency threshold is 80%, meaning that physical aisle areas with a user gaze focus frequency ≥ 80% are prime display areas. The three-dimensional coordinate codes of the aisles in the prime display area are columns 2-4 on the X-axis, layers 1-2 on the Y-axis, and column 1 on the Z-axis. The domain refers to the physical aisle area in the vending machine where the user's line of sight is most easily focused. It is obtained by the behavior data analysis module by statistically analyzing the user's line of sight focus position in historical video data. The three-dimensional coordinate code of spatially adjacent aisles refers to the code whose X-axis, Y-axis, or Z-axis coordinate values ​​differ from those of the aisle in the prime display area by 1. For example, if the coordinates of the aisle in the prime display area are X3-Y2-Z1, its adjacent coordinates are X2-Y2-Z1, X4-Y2-Z1, X3-Y1-Z1, etc. The remaining aisles are the aisles that are not occupied after the target aisle allocation for optimizing and adjusting the product set has been completed. When assigning the three-dimensional coordinate codes of the target aisles to other products, the constraint condition that the three-dimensional coordinate codes of the target aisles of the same category of products must be spatially adjacent or continuously arranged must be strictly followed.

[0042] In this embodiment, during the process of determining the target aisle 3D coordinate codes for products, a second precise screening is first performed on the high-frequency purchase product set and the high-attention, low-purchase product set to determine the core optimized product set and the core adjustment product set. Aisle resources are then preferentially allocated to these two types of core products. Next, using a pre-defined golden display area mapping table, the 3D coordinate codes of the golden display area aisles with the highest user attention frequency are assigned to products in the core optimized product set, and the 3D coordinate codes of the aisles adjacent to the golden display area are assigned to products in the core adjustment product set, ensuring the exposure rate of the two types of core products. Finally, for other products not included in the two core sets, the target aisle allocation is completed by combining the aisle adjustment constraints and the remaining aisle 3D coordinate codes of the vending machine, thus achieving the hierarchical allocation and rational configuration of vending machine aisle resources.

[0043] As can be seen from the above, this embodiment achieves hierarchical priority allocation of vending machine aisle resources. Core optimized products occupy the prime display area where users' eyes are most likely to focus, significantly increasing their product exposure and actual purchase rate. Core adjusted products are allocated to adjacent aisles in the prime display area, effectively solving the problem of their original hidden display positions and improving product conversion efficiency. Other products are allocated to the remaining aisles according to the aisle adjustment constraints, ensuring the rationality and standardization of the overall display layout of the vending machine and maximizing the utilization of vending machine aisle resources.

[0044] In one embodiment of this application, after generating the product display adjustment plan based on the association rule report, the method further includes: Step 1: Push the product display adjustment plan to the vending machine and related operating terminals; Step two: After the product display adjustment plan is implemented, collect new user transaction data and new user behavior data generated within the preset effect statistical period; Step 3: Based on the integrated data table, new user transaction data, and new user behavior data, determine the purchase increase rate of the product; Step 4: Implement a tiered iterative strategy based on the purchase increase rate; The implementation of a hierarchical iteration strategy includes: If the purchase increase rate is greater than or equal to the first preset increase threshold, the current product-sales channel correspondence will be used as the basis for data collection in the next statistical period. If the purchase increase rate is less than the first preset increase threshold but greater than or equal to the second preset increase threshold, then based on the new user transaction data and new user behavior data, the allocation of goods channels for other products in the product display adjustment plan, except for the high-frequency purchase product set, will be adjusted, and the process of steps one to four will be re-executed. If the purchase increase rate is less than the second preset increase threshold, then based on the new user transaction data and new user behavior data, the steps from data fusion to generating a new product display adjustment plan will be re-executed.

[0045] In this embodiment, the push method in step one can be 4G / 5G wireless communication. The vending machine receives the product display adjustment plan in JSON format, and the operation terminal receives the product display adjustment plan in encrypted PDF format. An SMS reminder is also sent to the operation terminal. A confirmation mechanism is in place after the push; if the cloud does not receive a confirmation signal from the vending machine or operation terminal within one minute, the plan is re-pushed. The operation terminal is the mobile phone, tablet, or other smart terminal of the vending machine operator. The effect statistics period in step two is consistent with the data collection statistics period, set to 7 days by default. The collection method and content of new user transaction data and new user behavior data are completely consistent with the collection method and content of the original data. After collection, a new first data table and a new second data table are generated respectively. The purchase increase rate in step three is the increase in the number of products after the display adjustment plan is implemented. The quantitative improvement indicator of purchase frequency relative to the previous period is calculated by comparing the merged data table with the newly collected user transaction data. The first preset improvement threshold is 10%, and the second preset improvement threshold is 5%. The hierarchical iteration strategy is to perform corresponding display optimization operations according to the different numerical ranges of the purchase improvement rate. The allocation of channels for other products except for the high-frequency purchase product set is adjusted. This means retaining the three-dimensional coordinate code of the target channel for high-frequency purchase products, and only making local fine-tuning to the target channels of core adjustment products and other products. The fine-tuning range is the adjacent channels of the original target channels. The steps from data fusion to generating a new product display adjustment plan are re-executed. This means that the new user transaction data and new user behavior data are fused, preprocessed, and association rule mining according to the original process to finally generate a brand-new product display adjustment plan to replace the original plan.

[0046] In this embodiment, after the product display adjustment plan is generated, it is pushed to both the vending machine and the operation terminal via 4G / 5G wireless communication to ensure that the plan is fully received. After the plan is executed on-site, an effect statistics cycle consistent with the original statistical cycle is started, and new user transaction data and behavior data are collected according to the original collection method. The newly collected transaction data is compared with the data in the original fused data table to calculate the product purchase increase rate. Based on the comparison results of the purchase increase rate with the first and second preset increase thresholds, the corresponding hierarchical iteration strategy is executed. If the plan is effective, it is retained and enters the next statistical cycle. If the plan effect is average, it is finely adjusted locally and the push and effect collection process is re-executed. If the plan is ineffective, a brand new plan is generated to form a complete closed-loop process for display optimization.

[0047] As can be seen from the above, this embodiment constructs a closed-loop system for the entire process of product display adjustment, from scheme generation and push to effect feedback and iterative optimization. It can promptly verify the actual operational effect of the display adjustment scheme, execute targeted layered iterative strategies based on the quantitative results of purchase increase rate, and avoid the drawbacks of unidirectional adjustment. It realizes dynamic and adaptive adjustment of vending machine product display, ensuring that the display optimization effect can be continuously improved. At the same time, the strategy of local fine-tuning reduces ineffective display adjustment operations and improves the overall efficiency of display optimization.

[0048] In one embodiment of this application, determining the purchase increase rate of a product includes: Calculate the first total number of purchases for the first category of products in the integrated data table within the statistical period, and determine the second total number of purchases for the first category of products within the performance statistical period based on the new user transaction data; The first category of goods is a collection of all goods after the merchandise display adjustment plan has been implemented; Based on the first and second total purchase counts, the purchase improvement rate of the first category of goods is calculated, which serves as the basis for implementing the tiered iterative strategy.

[0049] In this embodiment, the first category of goods is the set of all goods for sale in the vending machine after the goods display adjustment plan is implemented, without restrictions on the type or quantity of goods, ensuring the integrity of the data statistics scope; the first total purchase count is the sum of the purchase counts of all goods in the first category of goods in the original statistical period in the fusion data table, which is obtained by the data fusion preprocessing module extracting relevant data from the fusion data table and summarizing it; the second total purchase count is the sum of the purchase counts of all goods in the first category of goods in the new user transaction data collected within the effect statistical period, which is obtained by the sales data collection module collecting and classifying and summarizing it in real time; the specific calculation formula for the purchase improvement rate is: purchase improvement rate = (second total purchase count - first total purchase count) / first total purchase count × 100%, and the calculation result is rounded to two decimal places.

[0050] In this embodiment, when calculating the product purchase improvement rate used to determine the iterative strategy, the product set for data statistics is first clearly defined as all products on sale in the vending machine after the display adjustment plan is completed, to ensure the comprehensiveness and consistency of the purchase count statistics. Then, the purchase counts of all products in the product set within the original statistical period and the effect statistical period are extracted from the fused data table and the new user transaction data, respectively. The first total purchase count and the second total purchase count are then summarized. Finally, the overall purchase improvement rate of the product set is obtained by calculating the difference ratio, and this calculation result is used as the sole criterion for executing the hierarchical iterative strategy.

[0051] As can be seen from the above, this embodiment clearly defines the scope of the product set for calculating the purchase increase rate, avoiding calculation errors caused by fuzzy statistical boundaries, and ensuring that the calculation results of the purchase increase rate have a high degree of accuracy and objectivity. Using the overall purchase increase rate of all products sold in the vending machine as the basis for determining the iterative strategy ensures the uniformity of the determination criteria, provides reliable data support for the execution of the hierarchical iterative strategy, and ensures the scientific and rational nature of the display optimization iterative decision.

[0052] In one embodiment of this application, the effective dwell time and the number of times the gaze is focused are obtained by means of: Collect video data covering all product aisle areas of the vending machine; From the video data, identify the location where the user's gaze falls within the cargo aisle area; Based on the mapping relationship between the landing point position and the three-dimensional coordinate code, the target product that the user's line of sight is focused on is determined; For each target product, the cumulative time that all users' eyes continuously linger in its corresponding product aisle area within the statistical period is counted as the effective dwell time; The total number of times that all users' gaze effectively lingers in their corresponding cargo aisle area within the statistical period is counted as the number of gaze focuses.

[0053] In this embodiment, video data collection covering all product aisle areas of the vending machine can be accomplished by the vending machine's high-definition wide-angle camera module, installed centrally above the vending machine's control panel. Equipped with infrared illumination, it can clearly collect video data in low-light environments below 500 lux, with each frame containing a millisecond-level timestamp. The user's gaze focus is identified within the product aisle area from the video data, a process completed collaboratively by the behavior data analysis module using target detection and gaze tracking algorithms. The mapping relationship between the gaze focus and the 3D coordinate code is a pre-established one-to-one correspondence table, binding the physical location of the vending machine's product aisle area with the 3D coordinate code. After identifying the gaze focus, the corresponding 3D coordinate code can be directly matched. The target product is the product bound to the 3D coordinate code matched by the gaze focus location. The criterion for effective dwell time is that the user's gaze is continuously focused on the product aisle area corresponding to a target product for ≥1 second. The effective dwell time is the cumulative value of all dwell times meeting the effective dwell time criterion within the statistical period, in seconds. The number of gaze focuses is the cumulative value of all dwell times meeting the effective dwell time criterion within the statistical period, in times.

[0054] In this embodiment, a high-definition wide-angle camera module installed at a fixed position on the vending machine continuously collects video data covering all product aisle areas. The behavior data analysis module performs frame-level processing on the video data and identifies the user's gaze focus point within the product aisle area using target detection and gaze tracking algorithms. Then, through a pre-established mapping relationship between the gaze point position and the three-dimensional coordinate code, the target product corresponding to the gaze point position is matched. Subsequently, the duration of each user's gaze in the product aisle area corresponding to the target product within the statistical period is statistically analyzed. Valid gaze data is filtered out based on the ≥1 second valid gaze judgment standard. The duration of valid gaze and the number of valid gazes are accumulated and counted respectively, and finally the effective gaze duration and gaze focus count for each target product are obtained.

[0055] As can be seen from the above, this embodiment achieves full coverage acquisition of video data for all product aisle areas of the vending machine through a high-definition wide-angle camera module. Combined with target detection and eye-tracking algorithms, it accurately identifies the location of the user's gaze focus. The preset mapping relationship enables rapid matching of the gaze focus location to the target product. The effective dwell time ≥1 second criterion eliminates invalid gaze dwell data, ensuring the validity of behavioral data. The statistically obtained effective dwell time and gaze focus count can truly and accurately reflect the user's actual browsing habits, providing high-quality basic data for the calculation of the attention index and greatly improving the accuracy of user behavior data extraction.

[0056] In one embodiment of this application, identifying the location of the user's gaze focus within the cargo aisle area from video data includes: Detect and locate the user's facial region from each frame of the video data; Based on the facial region, locate the center position of the user's pupils; Based on the preset spatial geometric relationship model between the camera and the vending machine control panel, and the position of the pupil center, calculate the theoretical coordinates of the user's line of sight on the plane where the control panel is located; Identify the user's hand area in each frame of the image. When the user's hand is detected to be performing a selection operation on the operation panel, the theoretical landing point coordinates of the moment before the selection moment are used as the landing point of the user's gaze focus in the cargo channel area. The selection process includes tapping the touchscreen, pressing a physical button, or using gestures to select a specific area.

[0057] In this embodiment, the user's facial region is detected and located from each frame of the video data by the behavior data analysis module using the YOLOv5 target detection algorithm. This algorithm has a facial region recognition accuracy of ≥95%. In practical applications, other target detection algorithms can also be selected, and this embodiment does not limit this. The user's pupil center position is located based on the facial region by the behavior data analysis module using a gaze tracking algorithm based on pupil positioning. The spatial geometric relationship model between the camera and the vending machine's control panel is a mathematical model pre-established based on the camera's installation position, shooting angle, and the spatial positional relationship of the control panel, including the camera's focal length. Core parameters such as distance, installation height, and horizontal angle can be used to convert the pupil center position into theoretical landing point coordinates on the plane where the operation panel is located. The identification of the user's hand area in each frame of the image is also completed by the YOLOv5 object detection algorithm. The selection operation refers to the product selection operation performed by the user on the vending machine operation panel, specifically including clicking the product icon on the touch screen, pressing the product number corresponding to the physical button, and making a preset selection gesture in the gesture recognition area. The selection time is the millisecond-level timestamp corresponding to the user's selection operation, the previous time is the timestamp corresponding to the video data of the previous frame of the selection time, and the theoretical landing point coordinates are the coordinates calculated at the previous time.

[0058] In this embodiment, when identifying the user's gaze focus location from video data, the video data is first processed at the frame level. The user's facial region is detected and located in each frame image using a target detection algorithm. Then, the user's pupil center position is located based on the facial region using a gaze tracking algorithm. Combined with a pre-established spatial geometric relationship model, the pupil center position is converted into the theoretical coordinates of the user's gaze on the plane where the operation panel is located. At the same time, the user's hand region in each frame image is identified using a target detection algorithm, and whether the user is performing a selection operation on the operation panel is detected in real time. When a selection operation is detected, the theoretical coordinates of the moment before the selection time are selected as the actual location of the user's gaze focus in the cargo area, thus combining the user's gaze tracking with the actual purchase operation.

[0059] As can be seen from the above, this embodiment combines the theoretical landing point coordinates calculated by the eye-tracking algorithm with the user's actual purchasing operation, effectively correcting the landing point position deviation caused by pure algorithm calculation. This ensures that the identified eye focus landing point position highly matches the product that the user is actually paying attention to and preparing to purchase, significantly improving the accuracy of landing point position recognition. The user behavior data extracted based on the accurate landing point position is more in line with the user's actual browsing and purchasing habits, possessing high authenticity and reliability. This provides higher quality data support for subsequent attention index calculation, association rule mining, and display scheme generation, further improving the accuracy of product display optimization.

[0060] Based on the same principle as the user behavior-based product display method provided in the embodiments of this application, the embodiments of this application also provide a user behavior-based product display device, applied to a vending machine, the vending machine including multiple product channels, such as... Figure 2 As shown, the user behavior-based merchandise display device 20 may specifically include: a data acquisition module 21, a data fusion module 22, a feature extraction module 23, a report generation module 24, and a merchandise adjustment module 25. Data acquisition module 21 is used to acquire the first data table and the second data table of the vending machine. The first data table is a summary table of user transaction data for goods within the statistical period, and the second data table is a summary table of user behavior data generated when users select goods within the statistical period. Each product and multiple product channels in the vending machine are assigned a three-dimensional coordinate code. The data in the first data table and the second data table carry timestamps. The data fusion module 22 is used to merge the data in the first data table and the second data table, and to preprocess the merged data to obtain a merged data table. Feature extraction module 23 is used to determine the support, confidence and purchase conversion rate of various products based on the fused data table; The report generation module 24 is used to generate a report on the association rules between products and channels based on the support, confidence, and purchase conversion rate of various products, combined with the association rules between products and channels. The association rules are combinations of three-dimensional coordinate codes of products and channels that simultaneously meet the predefined support threshold, confidence threshold, and purchase conversion rate threshold. Product adjustment module 25 is used to generate product display adjustment plans based on association rule reports; For each product, The support level for each type of product is calculated based on the number of times the product is purchased in a specific product channel and the total number of times the product is purchased across all product channels. The confidence level of each type of product is calculated based on the number of times the product is purchased in a certain sales channel and the total number of times all products in that sales channel are purchased. The purchase conversion rate for each type of product is calculated based on the number of times the product is purchased and the product's attention index. The second data table includes at least the effective dwell time and the number of times the product's eyes are focused. The attention index is calculated based on the effective dwell time and the number of times the product's eyes are focused.

[0061] In one embodiment of this application, when generating a product display adjustment plan based on an association rule report, the product adjustment module 25 is specifically used for: Based on the association rule report, a set of high-frequency purchased products and a set of high-attention, low-purchase products are identified. The set of high-frequency purchased products consists of products with a purchase frequency higher than the first threshold, while the set of high-attention, low-purchase products consists of products with an attention index higher than the second threshold and a purchase conversion rate lower than the third threshold. Based on the set of frequently purchased goods, the set of high-attention but low-purchase goods, and the preset channel adjustment constraints, the three-dimensional coordinate code of the target channel for each goods is determined. Based on the original and target 3D coordinate codes of the products, a product display adjustment plan is generated. The product display adjustment plan includes the product code, the original 3D coordinate code, the target 3D coordinate code, and the basis for adjustment for each product. The constraints for adjusting the cargo lane include: For products belonging to the same category, the corresponding target channel three-dimensional coordinate codes are arranged adjacently or continuously in space.

[0062] In one embodiment of this application, the product adjustment module 25 is further used for: From the set of frequently purchased products, select the N products with the highest number of purchases to form the core optimized product set; From the set of high-attention, low-purchase products, select the top M products by attention index to form the core adjustment product set; Based on the preset golden display area mapping table, each product in the core optimized product set is assigned the three-dimensional coordinate code of the aisle corresponding to the golden display area as its target aisle three-dimensional coordinate code. The golden display area is the physical aisle area where the eye focus frequency is higher than the preset frequency threshold, which is determined by statistical analysis of historical user behavior data. For each item in the core product set, assign its target aisle three-dimensional coordinate code as the spatially adjacent aisle three-dimensional coordinate code of the aisle to the prime display area. For other goods not included in the core optimized goods set and the core adjusted goods set, a target three-dimensional coordinate code is assigned to them based on the lane adjustment constraints and the three-dimensional coordinate code of the remaining lanes.

[0063] In one embodiment of this application, the product display device 20 based on user behavior further includes: an optimization module, specifically used for: Step 1: Push the product display adjustment plan to the vending machine and related operating terminals; Step two: After the product display adjustment plan is implemented, collect new user transaction data and new user behavior data generated within the preset effect statistical period; Step 3: Based on the integrated data table, new user transaction data, and new user behavior data, determine the purchase increase rate of the product; Step 4: Implement a tiered iterative strategy based on the purchase increase rate; The implementation of a hierarchical iteration strategy includes: If the purchase increase rate is greater than or equal to the first preset increase threshold, the current product-sales channel correspondence will be used as the basis for data collection in the next statistical period. If the purchase increase rate is less than the first preset increase threshold but greater than or equal to the second preset increase threshold, then based on the new user transaction data and new user behavior data, the allocation of goods channels for other products in the product display adjustment plan, except for the high-frequency purchase product set, will be adjusted, and the process of steps one to four will be re-executed. If the purchase increase rate is less than the second preset increase threshold, then based on the new user transaction data and new user behavior data, the steps from data fusion to generating a new product display adjustment plan will be re-executed.

[0064] In one embodiment of this application, the optimization module is further configured to: Calculate the first total number of purchases for the first category of products in the integrated data table within the statistical period, and determine the second total number of purchases for the first category of products within the performance statistical period based on the new user transaction data; The first category of goods is a collection of all goods after the merchandise display adjustment plan has been implemented; Based on the first and second total purchase counts, the purchase improvement rate of the first category of goods is calculated, which serves as the basis for implementing the tiered iterative strategy.

[0065] In one embodiment of this application, the report generation module 24 is specifically used for: Collect video data covering all product aisle areas of the vending machine; From the video data, identify the location where the user's gaze falls within the cargo aisle area; Based on the mapping relationship between the landing point position and the three-dimensional coordinate code, the target product that the user's line of sight is focused on is determined; For each target product, the cumulative time that all users' eyes continuously linger in its corresponding product aisle area within the statistical period is counted as the effective dwell time; The total number of times that all users' gaze effectively lingers in their corresponding cargo aisle area within the statistical period is counted as the number of gaze focuses.

[0066] In one embodiment of this application, the report generation module 24 is further configured to: Detect and locate the user's facial region from each frame of the video data; Based on the facial region, locate the center position of the user's pupils; Based on the preset spatial geometric relationship model between the camera and the vending machine control panel, and the position of the pupil center, calculate the theoretical coordinates of the user's line of sight on the plane where the control panel is located; Identify the user's hand area in each frame of the image. When the user's hand is detected to be performing a selection operation on the operation panel, the theoretical landing point coordinates of the moment before the selection moment are used as the landing point of the user's gaze focus in the cargo channel area. The selection process includes tapping the touchscreen, pressing a physical button, or using gestures to select a specific area.

[0067] The apparatus in this application embodiment can execute the method provided in this application embodiment, and the implementation principle is similar. The actions performed by each module in the apparatus of each embodiment of this application correspond to the steps in the method of each embodiment of this application. For detailed functional descriptions of each module of the apparatus, please refer to the descriptions in the corresponding methods shown above, which will not be repeated here.

[0068] Figure 3 A schematic diagram of the structure of an electronic device to which this application embodiment applies is shown, such as... Figure 3 As shown, the electronic device can be used to implement the methods provided in any embodiment of this application.

[0069] like Figure 3 As shown, the electronic device 300 may primarily include at least one processor 301. Figure 3 The diagram shows components such as a memory 302, a communication module 303, and an input / output interface 304. Optionally, these components can be connected and communicate with each other via a bus 305. It should be noted that... Figure 3 The structure of the electronic device 300 shown is merely illustrative and does not constitute a limitation on the electronic devices to which the methods provided in the embodiments of this application are applicable.

[0070] The memory 302 can be used to store operating systems and applications, etc. The applications can include computer programs that implement the methods shown in the embodiments of this application when invoked by the processor 301, and can also include programs for implementing other functions or services. The memory 302 can be ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, RAM (Random Access Memory) or other types of dynamic storage devices that can store information and computer programs, or it can be EEPROM (Electrically Erasable Programmable Read Only Memory), CD-ROM (Compact Disc Read Only Memory) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer, but is not limited thereto.

[0071] Processor 301 is connected to memory 302 via bus 305 and implements corresponding functions by calling the application programs stored in memory 302. Processor 301 can be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 301 can also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0072] Electronic device 300 can connect to a network via communication module 303 (which may include, but is not limited to, components such as a network interface) to communicate with other devices (such as user terminals or servers) through the network and achieve data interaction, such as sending data to or receiving data from other devices. Communication module 303 may include wired network interfaces and / or wireless network interfaces, meaning the communication module may include at least one of wired or wireless communication modules.

[0073] The electronic device 300 can connect to necessary input / output devices, such as a keyboard or display device, via the input / output interface 304. The electronic device 300 itself may have a display device, and other display devices can also be connected externally via the interface 304. Optionally, a storage device, such as a hard drive, can also be connected via the interface 304, allowing data from the electronic device 300 to be stored, read, or transferred to the memory 302. It is understood that the input / output interface 304 can be a wired or wireless interface. Depending on the specific application scenario, the device connected to the input / output interface 304 can be an integral part of the electronic device 300 or an external device connected to the electronic device 300 when needed.

[0074] The bus 305 used to connect the components may include a path for transmitting information between the components. The bus 305 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. Depending on its function, the bus 305 may be divided into an address bus, a data bus, a control bus, etc.

[0075] Optionally, for the solution provided in the embodiments of this application, the memory 302 can be used to store a computer program that executes the solution of this application, and the processor 301 runs the computer program. When the processor 301 runs the computer program, it implements the operation of the method or apparatus provided in the embodiments of this application.

[0076] Based on the same principle as the method provided in the embodiments of this application, the embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, can implement the corresponding content of the aforementioned method embodiments.

[0077] This application also provides a computer program product, which includes a computer program that, when executed by a processor, can implement the corresponding content of the aforementioned method embodiments.

[0078] It should be noted that the terms "first," "second," "third," "fourth," "1," "2," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in a sequence other than that shown in the figures or text.

[0079] In the embodiments of this application, the terms "module" or "unit" refer to a computer program or part of a computer program that has a predetermined function and works with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0080] It should be understood that although arrows indicate various operation steps in the flowcharts of this application's embodiments, the order in which these steps are implemented is not limited to the order indicated by the arrows. Unless explicitly stated herein, in some implementation scenarios of this application's embodiments, the implementation steps in each flowchart can be executed in other orders as required. Furthermore, some or all steps in each flowchart, based on the actual implementation scenario, may include multiple sub-steps or multiple stages. Some or all of these sub-steps or stages can be executed at the same time, and each sub-step or stage can also be executed at different times. In scenarios where execution times differ, the execution order of these sub-steps or stages can be flexibly configured according to requirements, and this application's embodiments do not limit this.

[0081] The above description is only an optional implementation method for some implementation scenarios of this application. It should be noted that for those skilled in the art, other similar implementation methods based on the technical concept of this application without departing from the technical concept of this application also fall within the protection scope of the embodiments of this application.

Claims

1. A product display method based on user behavior, characterized in that, Applied to a vending machine, the vending machine including multiple product channels, the method includes: The vending machine acquires a first data table and a second data table. The first data table is a summary table of user transaction data for goods within a statistical period, and the second data table is a summary table of user behavior data generated when users select goods within a statistical period. Each of the goods and multiple product channels in the vending machine is assigned a three-dimensional coordinate code. The data in the first data table and the second data table carries a timestamp. The data in the first data table and the second data table are merged, and the merged data is preprocessed to obtain a merged data table; Based on the fused data table, the support, confidence, and purchase conversion rate of each type of product are determined. Based on the support, confidence, and purchase conversion rate of each type of product, and combined with the association rules between products and channels, an association rule report between products and channels is generated; the association rule is a combination of three-dimensional coordinate codes of products and channels that simultaneously satisfy predefined support thresholds, confidence thresholds, and purchase conversion rate thresholds. A product display adjustment plan is generated based on the aforementioned association rule report; For each product, The support level for each type of product is calculated based on the number of times the product is purchased in a specific product channel and the total number of times the product is purchased across all product channels. The confidence level of each type of product is calculated based on the number of times the product is purchased in a certain sales channel and the total number of times all products in that sales channel are purchased. The purchase conversion rate of each type of product is calculated based on the number of times the product is purchased and the attention index of the product. The second data table includes at least the effective dwell time and the number of times the eyes are focused. The attention index is calculated based on the effective dwell time and the number of times the eyes are focused.

2. The product display method based on user behavior as described in claim 1, characterized in that, The process of generating a product display adjustment plan based on the association rule report includes: Based on the association rule report, a set of frequently purchased products and a set of products with high attention but low purchases are identified; the set of frequently purchased products consists of products with a purchase frequency higher than a first threshold, and the set of products with high attention but low purchases consists of products with an attention index higher than a second threshold and a purchase conversion rate lower than a third threshold. Based on the set of frequently purchased goods, the set of high-attention but low-purchase goods, and the preset channel adjustment constraints, the target channel three-dimensional coordinate code of each of the goods is determined; Based on the original three-dimensional coordinate code of the product and the target three-dimensional coordinate code of the product, a product display adjustment plan is generated; the product display adjustment plan includes the product code, the original three-dimensional coordinate code of the product, the target three-dimensional coordinate code of the product, and the adjustment basis for each product; The cargo lane adjustment constraints include: For products belonging to the same category, the corresponding target channel three-dimensional coordinate codes are arranged adjacently or continuously in space.

3. The product display method based on user behavior as described in claim 2, characterized in that, The step of determining the target channel three-dimensional coordinate code for each product based on the set of frequently purchased products, the set of products with high attention but low purchases, and preset channel adjustment constraints includes: From the set of frequently purchased products, select the N products with the highest number of purchases to form the core optimized product set; From the set of high-attention, low-purchase products, the top M products in terms of attention index are selected to form the core adjustment product set; Based on the preset golden display area mapping table, each product in the core optimized product set is assigned a three-dimensional coordinate code of the aisle corresponding to the golden display area as its target aisle three-dimensional coordinate code. The golden display area is a physical aisle area where the eye focus frequency is higher than a preset frequency threshold, which is determined by statistical analysis of historical user behavior data. For each item in the core adjusted product set, assign its target aisle three-dimensional coordinate code as the aisle three-dimensional coordinate code of the aisle spatially adjacent to the prime display area. For other goods not included in the core optimized goods set and the core adjusted goods set, a target three-dimensional coordinate code is assigned to them based on the lane adjustment constraints and the three-dimensional coordinate code of the remaining lanes.

4. The product display method based on user behavior as described in claim 1, characterized in that, After generating the product display adjustment plan based on the association rule report, the method further includes: Step 1: Push the product display adjustment plan to the vending machine and associated operating terminal; Step 2: After the product display adjustment plan is implemented, new user transaction data and new user behavior data generated within the preset effect statistical period are collected. Step 3: Based on the fused data table, the new user transaction data, and the new user behavior data, determine the purchase increase rate of the product; Step 4: Based on the purchase increase rate, execute a tiered iterative strategy; The implementation of a hierarchical iteration strategy includes: If the purchase increase rate is greater than or equal to the first preset increase threshold, the current product-sales channel correspondence will be used as the basis for data collection in the next statistical cycle. If the purchase increase rate is less than the first preset increase threshold and greater than or equal to the second preset increase threshold, then based on the new user transaction data and the new user behavior data, the channel allocation of other products in the product display adjustment plan, except for the high-frequency purchase product set, will be adjusted, and the process of steps one to four will be re-executed. If the purchase increase rate is less than the second preset increase threshold, then based on the new user transaction data and the new user behavior data, the steps from data fusion to generating a new product display adjustment plan are re-executed.

5. The product display method based on user behavior as described in claim 4, characterized in that, Determining the purchase increase rate of the product includes: Calculate the first total number of purchases for the first category of products in the fused data table within the statistical period, and determine the second total number of purchases for the first category of products within the effect statistical period based on the new user transaction data; The first category of goods is a collection of all goods after the goods display adjustment plan has been implemented; Based on the first total number of purchases and the second total number of purchases, the purchase improvement rate of the first type of product set is calculated, which serves as the purchase improvement rate on which the tiered iterative strategy is based.

6. The product display method based on user behavior as described in claim 1, characterized in that, The methods for obtaining the effective dwell time and the number of times the gaze is focused include: Collect video data covering all product aisle areas of the vending machine; From the video data, identify the location where the user's gaze falls within the cargo aisle area; Based on the mapping relationship between the landing point position and the three-dimensional coordinate code, the target product that the user's line of sight is focused on is determined; For each target product, the cumulative time that all users’ gazes lingered continuously in its corresponding aisle area within the statistical period is counted as the effective dwell time; The total number of times that all users' gazes effectively lingered in their corresponding cargo aisle areas within the statistical period is counted as the number of gaze focuses.

7. The product display method based on user behavior as described in claim 6, characterized in that, The step of identifying the location of the user's gaze focus within the cargo aisle area from the video data includes: Detect and locate the user's facial region from each frame of the video data; Based on the facial region, locate the center position of the user's pupils; Based on the preset spatial geometric relationship model between the camera and the vending machine control panel, and the center position of the pupil, calculate the theoretical coordinates of the user's line of sight on the plane where the control panel is located; Identify the user's hand area in each frame of the image. When the user's hand is detected to be performing a selection operation on the operation panel, the theoretical landing point coordinates of the moment before the selection time corresponding to the operation are used as the landing point of the user's line of sight in the cargo channel area. The selection operation includes tapping the touchscreen, pressing a physical button, or making a selection in a specific area using gestures.

8. A merchandise display device based on user behavior, characterized in that, Applied to a vending machine, the vending machine including multiple product channels, the device includes: The data acquisition module is used to acquire a first data table and a second data table from the vending machine. The first data table is a summary table of user transaction data for goods within a statistical period, and the second data table is a summary table of user behavior data generated when users select goods within a statistical period. Each of the goods and multiple product channels in the vending machine is assigned a three-dimensional coordinate code. The data in the first data table and the second data table carries a timestamp. The data fusion module is used to merge the data in the first data table and the second data table, and to preprocess the merged data to obtain a merged data table. The feature extraction module is used to determine the support, confidence, and purchase conversion rate of various products based on the fused data table. The report generation module is used to generate a report on the association rules between products and channels based on the support, confidence, and purchase conversion rate of the various types of products, combined with the association rules between products and channels; the association rules are combinations of three-dimensional coordinate codes of products and channels that simultaneously satisfy predefined support thresholds, confidence thresholds, and purchase conversion rate thresholds. The product adjustment module is used to generate a product display adjustment plan based on the association rule report; For each product, The support level for each type of product is calculated based on the number of times the product is purchased in a specific product channel and the total number of times the product is purchased across all product channels. The confidence level of each type of product is calculated based on the number of times the product is purchased in a certain sales channel and the total number of times all products in that sales channel are purchased. The purchase conversion rate of each type of product is calculated based on the number of times the product is purchased and the attention index of the product. The second data table includes at least the effective dwell time and the number of times the eyes are focused. The attention index is calculated based on the effective dwell time and the number of times the eyes are focused.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing a computer program, and the processor executing the product display method based on user behavior as described in any one of claims 1 to 7 when running the computer program.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the product display method based on user behavior as described in any one of claims 1 to 7.

Images