Commodity interestingness analysis method based on user behaviors, computer equipment and storage medium

Abstract

The invention relates to the technical field of data analysis and processing, and discloses a commodity interestingness analysis method based on user behaviors, computer equipment and a storage medium, firstly, behaviors of users in a store are identified based on a computer vision technology, the limitation of dependence on sales data and customer feedback traditionally can be broken through, and the user experience is improved. Various behaviors and actions of the user in the purchasing process can be accurately captured, so that the potential interest of the user in the commodity can be more comprehensively known. And secondly, a more accurate and objective commodity interestingness report can be provided for merchants, and the merchants can be assisted to better perform commodity display, inventory management, marketing strategy formulation and the like, so that the sales conversion rate is improved, the market competitiveness of stores is enhanced, and greater economic benefits are brought to the merchants.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis and processing, and in particular to a method for analyzing product interest based on user behavior, a computer device, and a storage medium. Background Art

[0002] In traditional retail store operations, accurately grasping the level of user interest in various products is undoubtedly a core link of critical significance. The effective implementation of this link is of far-reaching significance and significant impact for merchants. It not only helps merchants to scientifically and rationally plan and adjust the layout of product displays, making the display of products more attractive and better in line with consumers' visual needs and shopping habits; it also plays an important role in inventory management, ensuring that the inventory level of products can not only meet market demand and avoid out-of-stock phenomena that affect sales performance, but also effectively control inventory costs and reduce capital occupation and losses caused by overstocked inventory; at the same time, for the formulation and implementation of marketing strategies, understanding the level of user interest in products provides a solid decision-making basis, allowing marketing activities to be targeted and improve marketing effectiveness and return on investment.

[0003] However, current analysis of product interest relies primarily on two traditional methods: sales data and customer feedback. Both of these analytical approaches suffer from varying degrees of limitations in practical application. Sales data only reflects the final stages of a user's purchase decision, but it struggles to effectively capture and analyze the underlying interest expressed throughout the entire shopping process. This means that sales data often presents a partial picture, failing to fully reflect a user's true interest in a product, potentially leading to biased product marketing decisions. While customer feedback can provide a degree of subjective evaluation and opinion, it is often highly subjective and susceptible to influences such as personal emotions, cognitive biases, and expressive abilities, resulting in less than objective and comprehensive results. Furthermore, the collection of customer feedback can be plagued by issues such as insufficient sample size and limited feedback channels, further limiting the accuracy and reliability of product interest analysis.

[0004] Therefore, there is an urgent need for a method that can analyze product interest more accurately and objectively.

[0005] The above information is presented as background information only to assist with an understanding of the present disclosure and is not a determination or admission that any of the above may be applicable as prior art with respect to the present disclosure. Summary of the Invention

[0006] The present invention provides a commodity interest analysis method based on user behavior, a computer device and a storage medium to solve the problems existing in the prior art.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] In a first aspect, the present invention provides a method for analyzing product interest based on user behavior, the method comprising:

[0009] S100: Obtaining surveillance video data of the store;

[0010] S200: Analyze the surveillance video data to identify user behaviors in the store; the behaviors include stopping to view products, taking products, and purchasing products;

[0011] S300: Determine the user interest in each product in the store based on the number of occurrences of each of the behavioral actions to obtain a product interest report.

[0012] Furthermore, in the method for analyzing product interest based on user behavior, S300 includes:

[0013] S301, assigning a corresponding weight value to each of the behaviors;

[0014] S302. Calculate the interest score of each product in the store based on the number of occurrences of each behavior action and its weight value, and comprehensively calculate the interest score of each behavior action corresponding to each product to determine the comprehensive user interest in each product to obtain a product interest report.

[0015] Furthermore, in the method for analyzing product interest based on user behavior, in step S301, the setting of the weight value satisfies the following conditions:

[0016] The weight value of the purchased product is greater than the weight value of the taken product;

[0017] The weight value of taking the product is greater than the weight value of staying to view the product.

[0018] Furthermore, in the method for analyzing product interest based on user behavior, after S300, the method further includes:

[0019] S401: Obtain product interest reports corresponding to different stores in the same area;

[0020] S402: Integrate the obtained product interest reports to identify highly competitive products and less competitive products; the highly competitive products are products that are viewed frequently and picked up frequently, while the less competitive products are products that are viewed frequently but picked up less frequently;

[0021] S403. Guiding the purchase of the highly competitive products to be increased and the purchase of the less competitive products to be reduced;

[0022] And / or, guide the allocation of high-competitiveness products to the best-selling area, and the allocation of low-competitiveness products to the non-best-selling area.

[0023] Furthermore, in the method for analyzing product interest based on user behavior, after S300, the method further includes:

[0024] S501: Obtain product interest reports corresponding to different stores in the same area;

[0025] S502: Integrate the obtained product interest reports;

[0026] S503: Input the integrated data into the basic large model for training to obtain a business assistant model that can provide accurate business suggestions.

[0027] Furthermore, in the method for analyzing product interest based on user behavior, after S300, the method further includes:

[0028] S701: Obtain product interest reports corresponding to different stores in the same area;

[0029] S702: Integrate the obtained product interest reports to form a regional product popularity profile;

[0030] S703: When a new user enters a regional store, recommending popular regional products to the new user based on the regional product popularity profile;

[0031] And / or, in combination with the geographical location of the new user, based on the regional product popularity portrait, popular regional products are pushed to the terminal device of the new user.

[0032] Furthermore, in the method for analyzing product interest based on user behavior, after S300, the method further includes:

[0033] S801: Obtain product interest reports corresponding to different stores in the same area;

[0034] S802: Integrate the obtained product interest reports to form a regional product popularity profile;

[0035] S803: Based on the regional product popularity profile, determine whether the region is suitable for opening a new store;

[0036] And / or, based on the regional product popularity portrait, recommend a popular product combination in the purchasing area to the new store.

[0037] Furthermore, in the method for analyzing product interest based on user behavior, after S300, the method further includes:

[0038] S601: Obtain product interest reports corresponding to different stores in the same area;

[0039] S602: Integrate the obtained product interest reports to form a regional product popularity profile;

[0040] S603: Combining the regional product popularity profile with the advertising campaign theme, a multimodal large model is used to dynamically generate and display popular product recommendation advertisements.

[0041] In a second aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, it implements the product interest analysis method based on user behavior as provided in the first aspect above.

[0042] In a third aspect, the present invention provides a computer-readable storage medium having computer-executable instructions stored thereon, wherein the computer-executable instructions are executed by a computer processor to implement the product interest analysis method based on user behavior as provided in the first aspect above.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] The present invention provides a method, computer device, and storage medium for analyzing product interest based on user behavior. First, by identifying user behavior within a store using computer vision technology, this method overcomes the limitations of traditional reliance on sales data and customer feedback, accurately capturing various user behaviors during the purchase process and thus providing a more comprehensive understanding of users' potential interest in a product. Second, it can provide merchants with more accurate and objective product interest reports, helping them better display products, manage inventory, and formulate marketing strategies, thereby increasing sales conversion rates, enhancing the store's market competitiveness, and generating greater economic benefits for merchants.

[0045] The present invention has other features and advantages that will be apparent from or will be described in detail in the accompanying drawings and the following detailed description incorporated herein, which together serve to explain certain principles of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0047] Figure 1 This is one of the flow charts of a method for analyzing product interest based on user behavior provided in the first embodiment of the present invention;

[0048] Figure 2 This is a second flow chart of a method for analyzing product interest based on user behavior provided in the first embodiment of the present invention;

[0049] Figure 3 This is a third flow chart of a method for analyzing product interest based on user behavior provided in the first embodiment of the present invention;

[0050] Figure 4 This is a fourth flow chart of a method for analyzing product interest based on user behavior provided in the first embodiment of the present invention;

[0051] Figure 5 This is the fifth flow chart of a method for analyzing product interest based on user behavior provided in the first embodiment of the present invention;

[0052] Figure 6 This is the sixth flow chart of a method for analyzing product interest based on user behavior provided in the first embodiment of the present invention;

[0053] Figure 7 This is the seventh flow chart of a method for analyzing product interest based on user behavior provided in the first embodiment of the present invention;

[0054] Figure 8 It is a structural diagram of a computer device provided in the second embodiment of the present invention. DETAILED DESCRIPTION

[0055] In order to explain in detail the possible application scenarios, technical principles, specific solutions that can be implemented, and the purpose and effects of this application, the following is a detailed description of the specific embodiments listed in conjunction with the accompanying drawings. The embodiments described herein are only used to more clearly illustrate the technical solutions of this application and are therefore only examples and are not intended to limit the scope of protection of this application.

[0056] References to "embodiments" herein mean that the specific features, structures, or characteristics described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the word "embodiment" in various places in the specification does not necessarily refer to the same embodiment, nor does it particularly limit its independence or relevance to other embodiments. In principle, in this application, as long as there are no technical contradictions or conflicts, the various technical features mentioned in the embodiments can be combined in any manner to form a corresponding implementable technical solution.

[0057] Unless otherwise defined, the technical terms used herein have the same meanings as those generally understood by those skilled in the art to which this application belongs; the use of relevant terms herein is only for describing specific embodiments and is not intended to limit this application.

[0058] In the description of this application, the term "and / or" is used to describe a logical relationship between objects, indicating that three relationships can exist. For example, A and / or B means: A exists, B exists, and both A and B exist. In addition, the character " / " in this document generally indicates that the objects before and after are in a logical "or" relationship.

[0059] In this application, terms such as "first" and "second" are merely used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual quantity, priority or sequence relationship between these entities or operations.

[0060] Without further limitations, in this application, the words "include", "comprise", "have" or other similar expressions used in the sentences are intended to cover non-exclusive inclusion. These expressions do not exclude the presence of additional elements in the process, method or product including the elements, so that the process, method or product including a series of elements may include not only those defined elements, but also other elements not explicitly listed, or elements inherent to such process, method or product.

[0061] In this application, expressions such as "greater than," "less than," and "exceed" are understood to exclude the number itself; expressions such as "above," "below," and "within" are understood to include the number itself. In addition, in the description of the embodiments of this application, "multiple" means two or more (including two), and similar expressions related to "multiple" are also understood in this way, such as "multiple groups" and "multiple times," unless otherwise specifically limited.

[0062] In the description of the embodiments of the present application, the space-related expressions used, such as "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "vertical", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise", "axial", "radial", "circumferential", etc., indicate the orientation or position relationship based on the orientation or position relationship shown in the specific embodiments or drawings, and are only for the convenience of describing the specific embodiments of the present application or facilitating the reader's understanding, and do not indicate or imply that the device or component referred to must have a specific position, a specific orientation, or be constructed or operated in a specific orientation. Therefore, it should not be understood as a limitation on the embodiments of the present application.

[0063] Unless otherwise expressly specified or limited, in the description of the embodiments of the present application, the terms "installed", "connected", "connected", "fixed", "set", etc. used should be understood in a broad sense. For example, the "connection" can be a fixed connection, a detachable connection, or an integrated setting; it can be a mechanical connection, an electrical connection, or a communication connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be the internal connection of two elements or the interaction relationship between two elements. For those skilled in the art of the present application, the specific meanings of the above terms in the embodiments of the present application can be understood according to the specific circumstances.

[0064] Example 1

[0065] Please refer to Figure 1 , Figure 1 This is a flow chart of a method for analyzing product interest based on user behavior, provided in Example 1 of the present invention. This method is applicable to scenarios where each product in a store is of interest to the user. The method can be implemented by software and / or hardware. The method specifically includes the following steps:

[0066] S100: Obtaining surveillance video data of a store.

[0067] It's important to note that this step is the data collection phase of the entire product interest analysis process. Store-wide surveillance equipment (such as cameras) collects in-store video information in real time or at scheduled intervals, and transmits this video data to a system or device for subsequent analysis.

[0068] Surveillance video data is the foundation for subsequent analysis of user behavior. It comprehensively records user activities at different times and locations within the store, containing rich information that can provide data support for accurate identification of user behavior.

[0069] Implementation: Utilize existing surveillance systems to transmit surveillance video data via the network to a local server or cloud storage platform. Ensure the integrity and clarity of the video data for subsequent analysis and processing.

[0070] S200: Analyze the surveillance video data to identify user behaviors in the store; the behaviors include stopping to view products, taking products, and purchasing products.

[0071] It should be noted that after acquiring surveillance video data, it is necessary to use computer vision and image processing technology to analyze the video data. By processing the video frames frame by frame and extracting features, the specific actions of the user in the video can be identified.

[0072] Accurately identifying user actions is key to understanding their level of interest in a product. Different actions reflect varying degrees of user interest and purchase intent. For example, stopping to view a product may indicate a certain level of interest; taking the product further indicates a deeper understanding or purchase intent; and purchasing the product directly reflects the conversion of interest into actual purchase behavior.

[0073] Implementation:

[0074] Computer vision technology: Object detection algorithms (such as YOLO and Faster R-CNN) can be used to detect the body outline and position of users in the video and determine the user's presence in the video.

[0075] Behavior recognition algorithms: These algorithms combine deep learning models (such as a combination of convolutional neural networks (CNNs) and recurrent neural networks (RNNs)) to classify and identify user actions. By training on a large amount of labeled user behavior data, the model can accurately distinguish between different behaviors, such as stopping to view a product, picking up a product, and purchasing it.

[0076] Feature extraction: Extracting user behavior features from videos, such as movement trajectory, dwell time, and movement amplitude, as a basis for behavior recognition. For example, analyzing how long a user dwells in front of a product can be used to determine how long they viewed the product.

[0077] S300: Determine the user interest in each product in the store based on the number of occurrences of each of the behavioral actions to obtain a product interest report.

[0078] It should be noted that after identifying user actions, the number of times each action occurs for each product is counted. Then, based on preset rules or algorithms, these number of action occurrences are converted into user interest indicators for the product, and a product interest report is generated.

[0079] By quantifying user behavior, we can objectively assess user interest in different products. Product interest reports provide merchants with intuitive and accurate data support, helping them understand the sales potential and market popularity of in-store products, thereby formulating more effective business strategies.

[0080] Implementation:

[0081] Action Counting: For each product, we count the number of actions such as stopping to view, picking up, and purchasing. By establishing a correlation between products and actions during video analysis, we can accurately attribute actions to the corresponding products.

[0082] Interest Calculation: Each action is assigned a weight based on its contribution to user interest. For example, a purchase action can be weighted higher because it ultimately reflects user interest, while a stop-and-watch action can be weighted lower. Then, based on the number of actions and their weight, a user interest score is calculated for each product. The calculation formula is: Interest Score = ∑ (Number of Actions × Action Weight).

[0083] Report Generation: The calculated user interest scores for each product are collated and analyzed to generate a product interest report. This report can be presented in tables, charts, and other formats, visually presenting each product's user interest ranking and detailed data. The report also provides analytical conclusions and recommendations to help merchants better understand user interest distribution and develop appropriate marketing strategies.

[0084] Through the above three steps, this method can accurately identify and analyze user behavior in stores, provide merchants with valuable product interest information, and help merchants improve their sales performance and market competitiveness.

[0085] Please refer to Figure 2 In one implementation of this embodiment, Figure 1 Specifically, the S400 is further refined, which specifically includes the following steps:

[0086] S301: assign a corresponding weight value to each of the behavioral actions.

[0087] It should be noted that the setting of weight values meets the following conditions:

[0088] The weight value of the purchased product is greater than the weight value of the taken product;

[0089] The weight value of taking the product is greater than the weight value of staying to view the product.

[0090] When analyzing product interest based on user behavior, different user actions reflect varying degrees of interest. Assigning a weight to each action quantifies this difference and clarifies the importance of each action in comprehensively assessing user interest in a product.

[0091] By assigning reasonable weights, we can more scientifically and accurately reflect the relationship between user behavior and product interest. For example, purchasing behavior is generally considered a direct reflection of a user's high interest in a product, and therefore should be assigned a higher weight. Meanwhile, simply stopping to look at a product briefly may be an unconscious action, and its significance as an indicator of product interest is relatively weak, so it should be assigned a lower weight.

[0092] Implementation:

[0093] Expert evaluation method: Invite industry experts, merchant representatives, and other experienced and professional personnel to evaluate different behaviors based on their experience and judgment, determine the relative importance of each behavior, and assign corresponding weights. For example, experts may assign a weight of 0.5 to purchasing, 0.3 to picking up an item, and 0.2 to stopping to view an item.

[0094] Data analysis: Collect a large amount of historical sales data and user behavior data, and use statistical analysis to identify correlations between different actions and actual product sales. Based on the strength of the correlation, assign a weight to each action. For example, if historical data shows that picking up a product significantly boosts final sales, the weight of this action can be increased accordingly.

[0095] Comprehensive method: Combining the results of expert evaluation and data analysis, the final weights are determined by comprehensively considering multiple factors. This method can fully utilize the advantages of expert experience and data analysis to improve the scientific nature and accuracy of weight allocation.

[0096] S302. Calculate the interest score of each product in the store based on the number of occurrences of each behavior action and its weight value, and comprehensively calculate the interest score of each behavior action corresponding to each product to determine the comprehensive user interest in each product to obtain a product interest report.

[0097] It's important to note that after assigning a weight to each action, the user interest score for each product is calculated based on the number of times each action occurs for each product. These scores are then combined to arrive at the overall user interest for each product. Finally, the overall user interest for all products is compiled into a report.

[0098] By calculating each product's interest score and overall user interest, we can intuitively quantify user interest in different products, providing merchants with clear and accurate data support. Product interest reports can help merchants understand the market popularity of in-store products, identify potential hot-selling and slow-moving items, and develop more effective product display, inventory management, and marketing strategies.

[0099] Implementation:

[0100] Calculate the interest score for each action: For each product, multiply the number of occurrences of each action by the weight of that action to obtain the product's interest score for that action. For example, if product A has 10 stop-and-watch events and a weight of 0.2, the interest score for product A for stop-and-watch events is 10 x 0.2 = 2 points.

[0101] Comprehensively calculate the overall user interest level for a product: Take the weighted sum of the interest scores for all actions corresponding to each product to obtain the overall user interest level for that product. For example, if the interest scores for product A for the actions of stopping to view, picking up, and purchasing are 2, 3, and 5, respectively, and the weights for these actions are 0.2, 0.3, and 0.5, respectively, the overall user interest level for product A is 2 × 0.2 + 3 × 0.3 + 5 × 0.5 = 3.8 points.

[0102] Generate a product interest report: Arrange the comprehensive user interest of all products in a certain order (such as from high to low) and generate a product interest report.

[0103] Through the above detailed steps, this implementation method can more scientifically and accurately evaluate the user's interest in the products in the store, providing merchants with more accurate and effective decision-making basis.

[0104] Please refer to Figure 3 In one implementation of this embodiment, Figure 1 The method is optimized based on the above, and the explanations of the terms that are the same or corresponding to the above embodiments are not repeated here. That is, after S300, the method further includes the following steps:

[0105] S401: Obtain the product interest reports corresponding to different stores in the same area.

[0106] It's important to note that this step focuses on different stores within the same area, aiming to comprehensively capture user interest data from every store within that area. By integrating this data, we can understand the characteristics and trends of demand for various products in the regional market at a macro level. For example, if there are multiple chain stores within a large commercial district, obtaining product interest reports for these stores can provide a glimpse into the product preferences of consumers across the entire district.

[0107] Reports can be automatically generated through the store's own sales management system and uploaded to the headquarters database, or a dedicated research team can regularly visit each store to collect relevant data and compile it into reports. These reports contain rich information, such as the number of times each product was viewed, picked up, and purchased in different stores, providing a data foundation for subsequent analysis.

[0108] S402. Integrate the obtained product interest reports to identify highly competitive products and less competitive products; the highly competitive products are products that are viewed more often and picked up more often, and the less competitive products are products that are viewed more often but picked up less often.

[0109] It's important to note that product interest reports from different stores are aggregated to create a unified data platform. On this platform, data can be categorized and organized by product category, brand, and other dimensions for comparative analysis. For example, snack data from all stores can be combined to analyze the performance of different snack styles and brands in each store.

[0110] Identification criteria for highly competitive products: Products that attract a high number of pauses, glances, and purchases. This indicates that the product attracts consumers' attention and, after capturing their attention, a high percentage of consumers further explore the product, demonstrating strong market competitiveness. For example, if a new flavor of puffed food attracts a large number of consumers to pause, glance, and try it out across multiple stores, then this puffed food can be considered a highly competitive product.

[0111] Low-competitiveness products: These products attract frequent viewing but are rarely picked up. While these products may attract consumers' attention, they may abandon their purchases as they learn more about the product due to factors such as taste, quality, or price. For example, a puffed food product with a pleasant flavor but a poor texture may attract many consumers to stop and view it, but due to dissatisfaction with the taste, few people pick it up. Therefore, this puffed food product can be considered a low-competitiveness product.

[0112] S403. Guiding the purchase of the highly competitive products to be increased and the purchase of the less competitive products to be reduced;

[0113] And / or, guide the allocation of high-competitiveness products to the best-selling area, and the allocation of low-competitiveness products to the non-best-selling area.

[0114] It's important to note that increasing procurement of highly competitive products: Based on identified highly competitive products, merchants can increase their purchases to meet market demand. Because highly competitive products are highly popular in the market, increasing procurement can ensure sufficient inventory supply, avoid stockouts, and further improve sales performance. For example, for the new flavors of puffed food mentioned above, merchants can negotiate with suppliers to increase order quantities to ensure they can meet consumer demand during peak sales season.

[0115] Reduce purchases of low-competitive items: For low-competitive items, merchants should appropriately reduce purchase volumes to avoid inventory overstocking. Since these items often sell poorly in the market, purchasing large quantities will increase inventory costs and tie up capital and storage space. For example, for a pair of shoes with a novel design but inferior material, merchants can reduce subsequent purchase orders or negotiate returns with suppliers to mitigate inventory risk.

[0116] Product Allocation Guidelines: Assign highly competitive products to best-selling areas: Assigning highly competitive products to the store's best-selling areas can increase product exposure and sales opportunities. These areas typically attract high traffic and consumer attention. Placing highly competitive products in these areas can attract more consumers and boost sales. For example, placing new flavors of puffed food prominently at the store entrance or setting up a dedicated best-selling product display area within the store can enhance product display effectiveness.

[0117] Assigning less competitive products to non-bestselling areas can reduce operating costs and avoid negatively impacting the sales of more competitive products. Non-bestselling areas have relatively low foot traffic and attract less consumer attention. Placing less competitive products in these areas can reduce the risk of inventory overstocking while freeing up display space for more competitive products. For example, a less-than-perfect puffed food product can be placed in a corner of the store or displayed alongside other slow-moving items.

[0118] Please refer to Figure 4 In one implementation of this embodiment, Figure 1 The method is optimized based on the above, and the explanations of the terms that are the same or corresponding to the above embodiments are not repeated here. That is, after S300, the method further includes the following steps:

[0119] S501: Obtain the product interest reports corresponding to different stores in the same area.

[0120] It's important to note that this step aims to comprehensively collect sales data from all stores within a region to understand consumer interest in various products across different stores. By integrating this data, we can understand the characteristics and trends of regional market demand at a macro level, providing data support for subsequent business decisions.

[0121] Relevant data can be obtained from the store's sales management system, membership system, cash register system, etc. These systems record consumer behavior information in the store, such as the length of time they stay to view products, the number of times they pick up products, purchase history, etc. By analyzing this data, a product interest report can be generated.

[0122] Collection scope: Clearly define the scope of the same area, such as a commercial district, different areas of a city, etc. Then collect product interest reports from all relevant stores in the area to ensure comprehensive and representative data.

[0123] Report content: A product interest report typically includes basic product information (such as name, category, brand, etc.), metrics such as the number of times a product is viewed, picked up, purchased, and purchase conversion rates at each store, as well as a product interest score calculated based on these metrics.

[0124] S502: Integrate the obtained product interest reports.

[0125] It's important to note that the purpose of integrating product interest reports from different stores is to eliminate data disparities and isolation, forming a unified, complete dataset. This allows for a more comprehensive and accurate analysis of product sales and consumer interest preferences across the entire region, providing a high-quality data foundation for subsequent model training.

[0126] Operation method:

[0127] Data cleaning: Data cleaning is performed on the obtained product interest reports to remove duplicate, erroneous, or incomplete data. For example, the data quality and accuracy are ensured by checking whether the product names are consistent and the data format is standardized.

[0128] Data alignment: Because data from different stores may differ, such as in statistical time and dimensions, data alignment is necessary. For example, the statistical time range can be unified, and data from different stores can be categorized and organized according to the same dimensions.

[0129] Data merging: Merge the cleaned and aligned data to form a large dataset containing product interest information from all stores. Data warehouses and databases can be used to store and manage this data, facilitating subsequent data analysis and model training.

[0130] S503: Input the integrated data into the basic large model for training to obtain a business assistant model that can provide accurate business suggestions.

[0131] It's important to note that by feeding the integrated product interest data into a large, basic model for training, the model learns the inherent patterns and relationships between product sales and consumer interest. The resulting business assistant model can provide precise business recommendations based on new product data and sales performance, helping merchants optimize product procurement, display, and marketing strategies, ultimately improving operational efficiency.

[0132] Operation method:

[0133] Select a basic large model: Choose an appropriate basic large model based on actual needs and data characteristics. For example, you can choose a pre-trained language model (such as BERT, GPT, etc.) as the basic model because these models have strong generalization and learning capabilities and can handle various types of data.

[0134] Model training: The integrated product interest data is used as a training set and fed into the underlying model for training. During training, the model's parameters and hyperparameters are adjusted to enable the model to accurately predict product sales and consumer interest preferences. Training methods such as supervised learning and reinforcement learning can be used to continuously optimize model performance based on training results.

[0135] Model evaluation and optimization: After training is complete, the model is evaluated to check whether indicators such as accuracy, recall, and F1 score meet the requirements. If the model performance is poor, the model can be optimized by adding training data, adjusting the model structure, and optimizing the training algorithm until the model reaches the expected performance level.

[0136] Application Deployment: Deploy the trained business assistant model into a real-world application environment and integrate it with the merchant's sales and management systems. Based on real-time product data and sales figures, the model can provide merchants with precise business recommendations, such as recommending best-selling products, optimizing product display locations, and developing personalized marketing strategies.

[0137] Please refer to Figure 5 In one implementation of this embodiment, Figure 1 The method is optimized based on the above, and the explanations of the terms that are the same or corresponding to the above embodiments are not repeated here. That is, after S300, the method further includes the following steps:

[0138] S701: Obtain the product interest reports corresponding to different stores in the same area.

[0139] S702: Integrate the obtained product interest reports to form a regional product popularity profile.

[0140] It's important to note that the purpose of integrating product interest reports from different stores is to eliminate data disparities and isolation, forming a unified, complete regional product popularity dataset. This allows for a more comprehensive and accurate analysis of product sales and consumer interest preferences across the entire region, providing precise data support for subsequent new user recommendations.

[0141] S703: When a new user enters a regional store, recommending popular regional products to the new user based on the regional product popularity profile;

[0142] And / or, in combination with the geographical location of the new user, based on the regional product popularity portrait, popular regional products are pushed to the terminal device of the new user.

[0143] It should be noted that when new users enter a regional store, popular regional products are recommended:

[0144] When a new user enters any store in the area, the store's identification system (such as facial recognition, membership card recognition, etc.) confirms the user's identity as a new user (not a user with existing consumption records) and obtains a regional product popularity profile.

[0145] Recommend relevant products to new users based on the regional product popularity profile, combined with the store's inventory and current promotions. For example, if the regional product popularity profile shows that a certain candy is selling well and popular in the region, and the store has stock and is currently promoting it, then this candy can be recommended to new users.

[0146] Push regional popular products based on the new user's geographic location:

[0147] When a new user enters a regional store, the geographic location information of their terminal device (such as a mobile phone) is obtained and combined with the regional product popularity portrait.

[0148] Based on the distance between the new user's location and each store, as well as the distribution of popular products in the area, information about popular products in the area can be pushed to the user's terminal device. For example, if a store near the new user's location has a promotion for a popular chewing gum, detailed information about the gum, a purchase link, and directions to nearby stores can be pushed to the user's terminal device.

[0149] It is understandable that information on popular regional products can be sent to the terminal devices of new users through mobile phone text messages, APP push notifications, etc.

[0150] Example:

[0151] For example, a supermarket chain has multiple stores in different regions. After obtaining product interest reports from each store in step S701, they integrate these reports to form a regional product popularity profile. When a new user enters a store in that region, the system, based on their geographic location (e.g., proximity to a popular shopping district) and the popularity of popular products in that district (e.g., popular snacks or seasonal fruits) in the regional product popularity profile, pushes relevant information to the user's terminal device, including product images, prices, promotions, and nearby store addresses, to guide the user toward purchase.

[0152] Please refer to Figure 6 In one implementation of this embodiment, Figure 1 The method is optimized based on the above, and the explanations of the terms that are the same or corresponding to the above embodiments are not repeated here. That is, after S300, the method further includes the following steps:

[0153] S801: Obtain the product interest reports corresponding to different stores in the same area.

[0154] S802: Integrate the obtained product interest reports to form a regional product popularity profile.

[0155] S803: Based on the regional product popularity profile, determine whether the region is suitable for opening a new store;

[0156] And / or, based on the regional product popularity portrait, recommend a popular product combination in the purchasing area to the new store.

[0157] It should be noted that to determine whether the area is suitable for opening a new store:

[0158] Based on regional product popularity profiles, analyze overall sales, popular product categories, and consumer demand trends. If sales of certain popular products in a region continue to grow, and consumer demand for specific categories is strong, this indicates significant market potential and a suitable location for new stores. Conversely, if product sales are poor and consumer demand is weak, it may not be suitable for new store openings.

[0159] Data analysis models and algorithms can be used to analyze and evaluate various indicators within regional product popularity profiles. For example, by calculating indicators such as average sales growth rate and market share, the level of competition and development potential in a region can be determined. Furthermore, by combining market research data with industry trends and comprehensively considering various factors, accurate judgments can be made.

[0160] Recommended popular product combinations in the purchasing area for new stores:

[0161] Based on regional product popularity profiles, we can understand consumers' preferences and demand trends for popular products in the region. Based on this information, we can recommend a selection of popular regional products for new stores to meet consumer demand and improve their market competitiveness.

[0162] Data analysis and machine learning algorithms can be used to determine the categories and quantities of popular products based on product sales data, consumer reviews, and other information from regional product popularity profiles. Furthermore, comprehensive analysis and optimization can be conducted based on factors such as the new store's positioning, target customer group, and inventory costs to generate personalized product mix recommendations. For example, if the regional product popularity profile shows high demand for electronics and disposable items among consumers in the area, and the new store is positioned as a young and fashionable consumer destination, a popular product mix such as Bluetooth headsets and disposable items could be recommended. Furthermore, recommendations can be dynamically adjusted and optimized based on factors such as product seasonality and fashion trends.

[0163] Example:

[0164] Suppose a chain retailer has multiple stores in different areas of a city. Through steps S801 and S802, it obtains and integrates product interest reports from each store to create a regional product popularity profile. This profile reveals that sales of electronic products (such as smartwatches and Bluetooth headsets) in a certain area are growing rapidly, and consumers have high evaluations of these products. Furthermore, this area has a high population density and high consumption potential. Combining market research data and industry trends, the company determines that this area is suitable for opening a new store. Furthermore, based on this regional product popularity profile, it recommends a selection of popular regional products, such as smartwatches and Bluetooth headsets, for the new store. The product mix is optimized based on the new store's positioning and target customer base to meet consumer demand and enhance market competitiveness.

[0165] Please refer to Figure 7 In one implementation of this embodiment, Figure 1 The method is optimized based on the above, and the explanations of the terms that are the same or corresponding to the above embodiments are not repeated here. That is, after S300, the method further includes the following steps:

[0166] S601: Obtain product interest reports corresponding to different stores in the same area;

[0167] S602: Integrate the obtained product interest reports to form a regional product popularity profile;

[0168] S603: Combining the regional product popularity profile with the advertising campaign theme, a multimodal large model is used to dynamically generate and display popular product recommendation advertisements.

[0169] It's important to note that the goal of this step is to generate popular product recommendation ads based on regional product popularity profiles and the current campaign theme, thereby increasing ad appeal and conversion rates. Multimodal large models can integrate information from multiple modalities, such as text, images, and audio, to generate richer, more vivid, and more accurate advertising content.

[0170] It's understandable that when integrating regional product popularity profiles with campaign themes, the regional product popularity profiles provide information on consumer interest trends and popularity distribution within the region, while the campaign themes clarify key elements such as the core content, target audience, and promotional focus of the advertising campaign. By combining these two, the resulting ads can be more targeted and effective. For example, if the campaign theme targets young people's fashion consumption, and the regional product popularity profiles indicate that a certain trendy sneaker is very popular among young people in the region, then this sneaker can be combined with the campaign theme to create an ad centered around elements such as fashion and trends.

[0171] Specifically, a multimodal big model is an AI model capable of processing data from multiple modalities (such as text, images, audio, and video). It understands and analyzes the connections and semantic information between different modal data, and uses this information for complex reasoning and creative creation. When generating popular product recommendation ads, the multimodal big model can automatically generate diverse ad content based on regional product popularity profiles and campaign themes, taking into account factors such as product characteristics, target audience preferences, and the characteristics of the advertising channel. For example, it can automatically generate advertising posters that align with the brand's image and advertising style based on product images and advertising copy. Alternatively, it can generate engaging audio ads based on product introductions and the cadence of the campaign.

[0172] In this embodiment, the ad generation process is not static but rather flexibly adjusted and optimized based on real-time data and user feedback. During ad generation, the multimodal large model can access the latest regional product popularity data and advertising performance feedback (such as click-through rate and conversion rate) in real time. Based on this information, it dynamically adjusts ad content, format, and delivery strategy. For example, if a product's popularity rapidly increases after an ad is released, the large model can automatically adjust the ad content to highlight the product's key selling points and increase its exposure. Furthermore, if a certain ad format is found to be ineffective among a specific user group, the large model can promptly switch to a different format to improve the ad's appeal and conversion rate.

[0173] When displaying the generated popular product recommendation ads, they can be presented to the target audience through appropriate channels and platforms. For example, display channels may include online platforms (such as e-commerce platforms, social media platforms, search engines, etc.) and offline channels (such as store electronic display screens, outdoor billboards, etc.). During the display process, it is necessary to optimize the display method, time, and frequency of ads based on the characteristics of different channels and user behavior habits to ensure that the ads can accurately reach the target audience and attract their attention and interest, thereby promoting product sales and promotion. For example, on social media platforms, ads can be accurately pushed to potential consumers based on users' interest tags and browsing history; on store electronic display screens, the content and frequency of ads can be dynamically adjusted based on customer traffic and customer characteristics at different time periods.

[0174] Although this application frequently uses terms such as interest, store, and product, the possibility of using other terms is not excluded. These terms are used only to more conveniently describe and explain the essence of the present invention; interpreting them as any additional limitations is contrary to the spirit of the present invention.

[0175] The present invention provides a method for analyzing product interest based on user behavior. First, by identifying user behavior within a store using computer vision technology, it can overcome the limitations of traditional reliance on sales data and customer feedback, accurately capturing various user behaviors during the purchase process, and thus more comprehensively understanding users' potential interest in products. Second, it can provide merchants with more accurate and objective product interest reports, helping them better display products, manage inventory, and formulate marketing strategies, thereby improving sales conversion rates, enhancing the store's market competitiveness, and generating greater economic benefits for merchants.

[0176] Example 2

[0177] Figure 8 A schematic diagram of the structure of a computer device provided in Embodiment 2 of the present invention. Figure 8 A block diagram of an exemplary computer device 12 suitable for use in implementing embodiments of the present invention is shown. Figure 8 The computer device 12 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.

[0178] like Figure 8 As shown, computer device 12 is implemented as a general-purpose computing device. Components of computer device 12 may include, but are not limited to, one or more processors or processing units 16, system memory 28, and a bus 18 that connects various system components (including system memory 28 and processing unit 16).

[0179] Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor, or a local bus using any of a variety of bus architectures. Examples of these architectures include, but are not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MAC) bus, an Enhanced ISA bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus.

[0180] The computer device 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the computer device 12, including volatile and non-volatile media, removable and non-removable media.

[0181] System memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. Computer device 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, storage system 34 may be configured to read and write non-removable, non-volatile magnetic media ( Figure 8 Not shown, usually called a "hard drive"). Although Figure 8 Although not shown, a magnetic disk drive for reading and writing to a removable non-volatile magnetic disk (e.g., a "floppy disk"), as well as an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to bus 18 via one or more data media interfaces. Memory 28 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of various embodiments of the present invention.

[0182] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in memory 28. Such program modules 42 include, but are not limited to, an operating system, one or more application programs, other program modules, and program data, each of which, or some combination thereof, may include an implementation of a network environment. Program modules 42 generally implement the functions and / or methodologies of the embodiments described herein.

[0183] The computer device 12 may also communicate with one or more external devices 14 (e.g., a keyboard, a pointing device, a display 24, etc.), one or more devices that enable a user to interact with the computer device 12, and / or any device that enables the computer device 12 to communicate with one or more other computing devices (e.g., a network card, a modem, etc.). Such communication may be performed via an input / output (I / O) interface 22. Furthermore, the computer device 12 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 20. As shown, the network adapter 20 communicates with the other modules of the computer device 12 via the bus 18. It should be understood that although Figure 8 Not shown, other hardware and / or software modules may be used in conjunction with computer device 12, including but not limited to microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0184] The processing unit 16 executes various functional applications and data processing by running programs stored in the system memory 28 , such as implementing the commodity interest analysis method based on user behavior provided in an embodiment of the present invention.

[0185] Example 3

[0186] A third embodiment of the present invention provides a computer-readable storage medium having computer-executable instructions stored thereon. When the instructions are executed by a processor, the method for analyzing product interest based on user behavior as provided in all the embodiments of the present application is implemented.

[0187] Any combination of one or more computer-readable media may be employed. A computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an 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 this document, 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, apparatus, or device.

[0188] A computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. Such propagated data signals may take various 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.

[0189] Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0190] Computer program code for performing the operations of the present invention may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0191] Finally, it should be noted that although the above embodiments have been described in the specification and drawings of this application, this does not limit the scope of patent protection of this application. All technical solutions generated by replacing or modifying equivalent structures or equivalent processes based on the essential concepts of this application using the contents recorded in the specification and drawings of this application, as well as directly or indirectly implementing the technical solutions of the above embodiments in other related technical fields, are included in the scope of patent protection of this application.

Claims

1. A method for analyzing product interest based on user behavior, characterized in that: The method comprises: S100: Obtaining surveillance video data of the store; S200: Analyze the surveillance video data to identify user behaviors in the store; the behaviors include stopping to view products, taking products, and purchasing products; S300: Determine the user interest in each product in the store based on the number of occurrences of each of the behavioral actions to obtain a product interest report.

2. The method for analyzing product interest based on user behavior according to claim 1, characterized in that: The S300 includes: S301, assigning a corresponding weight value to each of the behaviors; S302. Calculate the interest score of each product in the store based on the number of occurrences of each behavior action and its weight value, and comprehensively calculate the interest score of each behavior action corresponding to each product to determine the comprehensive user interest in each product to obtain a product interest report.

3. The method for analyzing product interest based on user behavior according to claim 2, characterized in that: In S301, the weight value is set to meet the following conditions: The weight value of the purchased product is greater than the weight value of the taken product; The weight value of taking the product is greater than the weight value of staying to view the product.

4. The method for analyzing product interest based on user behavior according to claim 1, characterized in that: After S300, the method further includes: S401: Obtain product interest reports corresponding to different stores in the same area; S402: Integrate the obtained product interest reports to identify highly competitive products and less competitive products; the highly competitive products are products that are viewed frequently and picked up frequently, while the less competitive products are products that are viewed frequently but picked up less frequently; S403. Guiding the purchase of the highly competitive products to be increased and the purchase of the less competitive products to be reduced; And / or, guide the allocation of high-competitiveness products to the best-selling area, and the allocation of low-competitiveness products to the non-best-selling area.

5. The method for analyzing product interest based on user behavior according to claim 1, characterized in that: After S300, the method further includes: S501: Obtain product interest reports corresponding to different stores in the same area; S502: Integrate the obtained product interest reports; S503: Input the integrated data into the basic large model for training to obtain a business assistant model that can provide accurate business suggestions.

6. The method for analyzing product interest based on user behavior according to claim 1, characterized in that: After S300, the method further includes: S701: Obtain product interest reports corresponding to different stores in the same area; S702: Integrate the obtained product interest reports to form a regional product popularity profile; S703: When a new user enters a regional store, recommending popular regional products to the new user based on the regional product popularity profile; And / or, in combination with the geographical location of the new user, based on the regional product popularity portrait, popular regional products are pushed to the terminal device of the new user.

7. The method for analyzing product interest based on user behavior according to claim 1, characterized in that: After S300, the method further includes: S801: Obtain product interest reports corresponding to different stores in the same area; S802: Integrate the obtained product interest reports to form a regional product popularity profile; S803: Based on the regional product popularity profile, determine whether the region is suitable for opening a new store; And / or, based on the regional product popularity portrait, recommend a popular product combination in the purchasing area to the new store.

8. The method for analyzing product interest based on user behavior according to claim 1, characterized in that: After S300, the method further includes: S601: Obtain product interest reports corresponding to different stores in the same area; S602: Integrate the obtained product interest reports to form a regional product popularity profile; S603: Combining the regional product popularity profile with the advertising campaign theme, a multimodal large model is used to dynamically generate and display popular product recommendation advertisements.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the method for analyzing product interest based on user behavior according to any one of claims 1 to 8 is implemented.

10. A computer-readable storage medium having computer-executable instructions stored thereon, characterized in that: The computer executable instructions are executed by a computer processor to implement the commodity interest analysis method based on user behavior according to any one of claims 1 to 8.

Images