Collection-oriented advertisement pushing method, device, equipment, medium and product

Abstract

The embodiment of the invention provides a collective-oriented advertisement pushing method and device, equipment, a medium and a product, and the method comprises the steps: obtaining a feature data set of a target collective through analyzing the communication data of a network access service shared by a plurality of collective members of the target collective, and carrying out the advertisement pushing of the target collective based on an active time period and a commodity interest set corresponding to each collective member, calculating a target collective commodity interest set corresponding to the current time, selecting commodities meeting an interestingness threshold value as recommended commodities based on the target collective commodity interest set, then obtaining feature data of a plurality of shops, calculating a comprehensive score of each shop based on the recommended commodities and the feature data of the plurality of shops, and obtaining the recommended commodities; and selecting the shops with the comprehensive scores meeting the score threshold as the target shops, and finally generating personalized advertisements based on the target collections, the recommended commodities and the target shops by applying an advertisement generation model, and pushing the personalized advertisements to the target collections, so that the shops can be accurately pushed to the target customer collections.

Description

Technical Field

[0001] This application belongs to the field of information recommendation technology, and in particular relates to a method, apparatus, device, medium and product for collective advertising push. Background Technology

[0002] Street-front shops are typically retail stores located on both sides of a street, facing nearby community members. The rapid growth of street-front shops has played a positive role in promoting economic and social development, contributing significantly to job creation, economic growth, and urban beautification, while also providing convenient services to the community.

[0003] In existing technologies, the main approach is to analyze the behavior of individual group members to recommend shops or products of interest to mobile users, or to match nearby shops to group members based on their location information.

[0004] However, existing technologies do not take into account the mutual influence between group members, so how to accurately push these shops to the target customer group is an urgent problem to be solved. Summary of the Invention

[0005] This application provides a method, apparatus, device, medium, and product for targeted advertising to a group, which can accurately push shops to a target customer group.

[0006] In a first aspect, embodiments of this application provide a method for pushing advertisements to a group, the method comprising: By analyzing the communication data of multiple group members sharing network access services of the target group, a set of characteristic data of the target group is obtained. The set of characteristic data includes the characteristic data of multiple group members. The characteristic data of each group member includes: group member identifier, active time period and product interest set. Based on the active time period and product interest set corresponding to each group member, calculate the target group product interest set corresponding to the current time, and select products that meet the interest threshold as recommended products based on the target group product interest set. The target group product interest set includes: product identifier and the interest level corresponding to the product identifier. Acquire feature data for multiple shops. The feature data for each shop includes: product list, positive feature data that influences the selection of shops by group members, and negative feature data. Based on the feature data of recommended products and multiple shops, calculate the comprehensive score of each shop, and select the shops whose comprehensive scores meet the score threshold as target shops. Based on the target audience, recommended products, and target stores, an advertising generation model is used to generate personalized ads and push them to the target audience.

[0007] In one optional implementation of the first aspect, based on the active time period and product interest set corresponding to each group member, the target group product interest set corresponding to the current time is calculated, including: Based on the active time period corresponding to each group member, a first feature data set corresponding to the current time is selected. The first feature data set includes the first group member feature data of multiple group members. Based on the set of product interests in the feature data of the first group member corresponding to each group member, calculate the interest of each group member in each product and the purchase intention of each group member; Based on each group member's interest in each product and their willingness to purchase, the target group's interest in each product is calculated, and a target group product interest set is constructed based on the target group's interest in each product.

[0008] In one optional implementation of the first aspect, the product interest set corresponding to each collective member includes multiple product interests, and each product interest includes: purchase frequency and interest behavior data; Based on the product interest set in the first group member feature data corresponding to each group member, the interest level of each group member in each product and the purchase intention of each group member are calculated, including: Based on interest behavior data, calculate the level of interest of each group member in each product; Based on purchase frequency and interest behavior data, the purchase intention of each group member is calculated.

[0009] In one alternative implementation of the first aspect, the purchase intention of each group member is calculated based on purchase frequency and interest behavior data, including: Based on the product interest set corresponding to each group member, obtain the total number of purchases of all products in the product interest set within a preset time window, as well as the total number of occurrences of interest behavior data for all products in the product interest set; The ratio of the total number of purchases to the total number of occurrences of interest behavior data is calculated as the purchase intention of each group member.

[0010] In an optional implementation of the first aspect, based on the feature data of recommended products and multiple shops, a comprehensive score is calculated for each shop, and shops whose comprehensive scores meet a score threshold are selected as target shops, including: Based on the recommended products and the product list corresponding to each store, select stores to be recommended. The product list corresponding to the stores to be recommended contains the recommended products. For each recommended store, the positive and negative feature data are standardized to obtain dimensionless positive and negative feature data. The first function is selected to normalize the dimensionless positive feature data to obtain normalized positive feature data. The first function is configured such that the normalized positive feature data increases as the dimensionless positive feature data increases. The second function is selected to normalize the negative feature data that eliminates the dimension, resulting in normalized negative feature data. The second function is configured such that the normalized negative feature data decreases as the negative feature data that eliminates the dimension increases. For each shop to be recommended, a comprehensive score is calculated based on normalized positive feature data, normalized negative feature data, and their respective weights.

[0011] In one optional implementation of the first aspect, a personalized advertisement is generated based on the target audience, recommended products, and target stores using an advertising generation model, and then pushed to the target audience, including: Get a preset advertising information template, which contains structured fields for generating advertising copy; The target audience, recommended products, and characteristic data of the target store are used as variables and input into a preset advertising information template to generate advertising information. The advertising information is input into the advertising generation model to generate personalized ads that match the target store and the target group. The advertising generation model is a generative artificial intelligence model. Personalized ads are pushed to the display terminals corresponding to the target group.

[0012] Secondly, embodiments of this application provide a group-oriented advertising push device, the device comprising: The first acquisition module is used to obtain a set of characteristic data of the target group by parsing the communication data of multiple group members sharing the network access service. The set of characteristic data includes the characteristic data of multiple group members. Each group member's characteristic data includes: group member identifier, active time period, and set of product interests. The first calculation module is used to calculate the target group's product interest set corresponding to the current time based on the active time period and product interest set of each group member, and select products that meet the interest threshold as recommended products based on the target group's product interest set. The target group's product interest set includes: product identifier and the interest level corresponding to the product identifier. The second acquisition module is used to acquire feature data of multiple shops. The feature data of each shop includes: a product list, positive feature data that influences the selection of shops by group members, and negative feature data. The second calculation module is used to calculate the comprehensive score of each store based on the feature data of recommended products and multiple stores, and select stores whose comprehensive scores meet the score threshold as target stores. The push module is used to generate personalized ads based on the target audience, recommended products, and target stores, and then push them to the target audience.

[0013] Thirdly, embodiments of this application provide an electronic device, the device including: a processor, and a memory storing computer program instructions; The processor reads and executes computer program instructions to implement a collective advertising push method for any of the first aspects.

[0014] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer program instructions, which, when executed by a processor, implement a collective advertising push method as described in any of the first aspects.

[0015] Fifthly, embodiments of this application provide a computer program product, characterized in that, when the instructions in the computer program product are executed by the processor of an electronic device, the electronic device performs a collective advertising push method as described in the first aspect.

[0016] The application provides a method, apparatus, device, medium, and product for pushing advertisements to a group. It can obtain a set of characteristic data for the target group by parsing the communication data of multiple group members sharing network access services. Based on the active time period and product interest set corresponding to each group member, it calculates the target group's product interest set for the current time. Based on the target group's product interest set, it selects products that meet an interest threshold as recommended products. Then, it obtains characteristic data from multiple shops, calculates a comprehensive score for each shop based on the recommended products and the characteristic data of multiple shops, and selects shops whose comprehensive scores meet a score threshold as target shops. Finally, based on the target group, recommended products, and target shops, it applies an advertising generation model to generate personalized advertisements and pushes them to the target group. Compared to existing technologies that only analyze the behavior of individual group members to recommend shops or products, this application deeply integrates the active time periods and product interests of all group members through shared network communication data of the target group. It generates a group-level product interest set that fits the current time period through quantitative calculation. When matching shops, it takes the overall needs of the group as the core, combines the positive and negative feature data of the shops to calculate the comprehensive score of the shops and select target shops. Finally, the personalized advertisements generated based on the target group, recommended products and target shops are more in line with the common needs and scenario characteristics of the group. After being pushed to the target group, it can achieve the dissemination effect of one person paying attention and the whole family being reached, thereby significantly improving the reach and exposure rate of shop advertisements. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 A flowchart illustrating a collective advertising push method according to an embodiment of this application is shown. Figure 2 A flowchart illustrating a collective advertising push method according to another embodiment of this application is shown; Figure 3 A flowchart illustrating a collective advertising push method according to another embodiment of this application is shown; Figure 4 A flowchart illustrating a collective advertising push method according to another embodiment of this application is shown; Figure 5 This is an overall architecture diagram of a group-oriented advertising push system; Figure 6 This is a schematic diagram of the overall process of an advertising push system for a group. Figure 7 This is a schematic diagram of the structure of an advertising push device for a group provided in this application; Figure 8 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0019] The features and exemplary embodiments of various aspects of this application will be described in detail below. To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain this application and not to limit it. For those skilled in the art, this application can be implemented without some of these specific details. The following description of the embodiments is merely to provide a better understanding of this application by illustrating examples.

[0020] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0021] It should be noted that the acquisition, storage, use, and processing of data in this application embodiment all comply with the relevant provisions of national laws and regulations.

[0022] First, let me explain the terms used in this application: Group member profile: A group member profile consists of static group member attributes such as gender, age, and region, and dynamic group member attributes such as interests and consumption behavior. Essentially, it is a tagged group member information that uses actual data to outline the profile of the target group members, thereby helping to accurately understand the target group members.

[0023] Family profile: A collection built upon the profiles of each member within a family group.

[0024] Generative Artificial Intelligence (AIGC) refers to artificial intelligence techniques based on generative adversarial networks, large-scale pre-trained models, and other methods. It utilizes existing data for learning and recognition to generate relevant content with appropriate generalization capabilities. The core idea of ​​AIGC is to use artificial intelligence algorithms to generate content with a certain degree of creativity and quality. Through training models and learning from large amounts of data, AIGC can generate relevant content based on input conditions or guidance. For example, by inputting keywords, descriptions, or samples, AIGC can generate matching articles, images, audio, etc.

[0025] Currently, methods primarily analyze the behavior of individual group members to recommend shops or products they might be interested in, or recommend nearby shops based on their location information. However, existing technologies do not address the mutual influence between group members. For example, in groups such as families, colleagues, or friends, the interests, preferences, and consumption decisions among group members often exhibit significant correlations and synergies. Individual decisions are not only influenced by their own preferences but are also frequently affected by the needs, recommendations, or shared behavioral patterns of other members within the group. Therefore, how to effectively push shops to target customer groups is a problem that urgently needs to be solved.

[0026] Based on this, this application provides a method for targeted advertising to groups. Starting from the group dimension, it obtains the characteristic data set of the target group by parsing the communication data of multiple group members sharing network access services, and deeply explores the common interests and real-time active time periods of group members. By obtaining the characteristic data of shops, it analyzes the matching degree between shops and group interests in real time, and dynamically generates personalized advertising content based on multi-dimensional evaluation indicators. Finally, it accurately pushes advertisements during the active time periods of group members, thereby achieving precise advertising to groups and significantly improving the reach and exposure rate of shop advertisements.

[0027] To address the problems of the prior art, embodiments of this application provide a method, apparatus, device, medium, and product for collective advertising push.

[0028] The following section first introduces a method for pushing advertisements to a group, as provided in the embodiments of this application.

[0029] Figure 1 This illustration shows a flowchart of a group-oriented advertising push method according to an embodiment of this application. Figure 1 As shown, the method may include the following steps: S101: By parsing the communication data of multiple group members sharing the network access service of the target group, the characteristic data set of the target group is obtained. The characteristic data set includes the characteristic data of multiple group members. The characteristic data of each group member includes: group member identifier, active time period and product interest set.

[0030] In this embodiment of the application, by parsing the communication data of multiple group members of the target group sharing the network access service, a set of characteristic data of the target group is obtained. The set of characteristic data of the target group includes the characteristic data of multiple group members. Each group member's characteristic data includes: group member identifier, active time period, and product interest set.

[0031] In one example, based on the internet traffic of group members under fixed broadband, Deep Packet Inspection (DPI) technology can be used to deeply analyze and reconstruct the application layer information in the seven-layer protocol of Open Systems Interconnection (OSI) by deeply reading the content of Internet Protocol Packet (IP packet) payload. This can yield complete application content and extract the group member characteristic data of multiple group members.

[0032] In one example, if the target group is a family group, then a complete family profile can be constructed by combining the complete internet access data of the group members owned by the operator with the actual consumption data of the group members provided by the e-commerce platform. This profile represents the characteristic data set of the target group. Target groups can also include groups sharing networks, such as businesses, dormitories, or communities.

[0033] The following describes a specific implementation method for obtaining a feature dataset of a target group, using a family group as an example: First, family member tags are created to describe the attributes of each member, including static attributes (age, gender), active periods, and group member behaviors (browsing history, purchase history, adding items to cart), etc. Then, by parsing the home broadband uplink data, the mobile phone numbers of resident group members under the home broadband account are analyzed. Family member information is identified through mobile phone numbers. Under the home broadband account, each mobile phone number corresponds to one group member; therefore, data is first obtained according to the family member tags created above, thus unifying the data generated by each group member.

[0034] For example, when authorized by the group members, their age and gender can be obtained from the group member information database, or their gender and age range can be inferred from their browsing and search behavior; the active time periods of each family member at home can be analyzed by analyzing their mobile phone broadband connection time; and products and services that the group members are interested in can be mined from their purchase, shopping cart, and web browsing history. Table 1 shows the characteristic data set of the family group.

[0035]

[0036] S102: Based on the active time period and product interest set corresponding to each group member, calculate the target group product interest set corresponding to the current time, and select products that meet the interest threshold as recommended products based on the target group product interest set. The target group product interest set includes: product identifier and the interest level corresponding to the product identifier.

[0037] In this embodiment, the active time periods of group members differ. For example, for a family group, parents are only at home in the evening. If the product interest set is directly merged without considering the time period, it will cause the target group's product interest set to become disconnected from the actual needs of the current target group. Therefore, it is necessary to dynamically calculate the target group's product interest set in combination with the current time, and select products that meet the interest threshold as recommended products based on the target group's product interest set.

[0038] In one example, if the current time is 7 o'clock, the product interest set corresponding to member number 1 and the product interest set corresponding to member number 2 need to be selected together to calculate the target collective product interest set; if the current time is 18 o'clock, only the product interest set corresponding to member number 2 needs to be selected to calculate the target collective product interest set.

[0039] S103: Obtain feature data for multiple shops. The feature data for each shop includes: a product list, positive feature data that influences the selection of shops by group members, and negative feature data. In this embodiment, feature data of multiple shops is obtained. In one example, feature data of multiple shops can be obtained through data provided during registration in the shop system. The feature data of each shop includes: a product list, positive feature data influencing group members' shop selection, and negative feature data. In one example, the positive feature data influencing group members' shop selection includes: shop rating and product sales volume; the negative feature data influencing group members' shop selection includes: product price, shop distance, and acquisition time, where acquisition time refers to the time spent acquiring products from the shop, including self-pickup time and delivery time. In one example, when the positive or negative feature data is dynamically changing, the average value of the feature data within a preset time period is taken as the corresponding positive or negative feature data. For example, the average value of shop ratings over the past 7 days is selected as the positive feature data.

[0040] S104: Based on the feature data of recommended products and multiple shops, calculate the comprehensive score of each shop, and select the shops whose comprehensive scores meet the score threshold as target shops.

[0041] In this embodiment, based on the recommended products and the product list of each store, stores whose product lists contain the recommended products are selected as potential recommended stores. Then, based on the positive and negative feature data of the potential recommended stores, a comprehensive score is calculated for each store, and stores whose comprehensive scores meet the score threshold are selected as target stores. In one example, the store with the highest comprehensive score can be selected as the target store.

[0042] S105: Based on the target audience, recommended products, and target stores, apply the advertising generation model to generate personalized ads and push them to the target audience.

[0043] In this embodiment of the application, an advertising generation model is applied, and corresponding personalized advertisements are generated based on the target group, recommended products, and target stores, and the generated personalized advertisements are pushed to the target group.

[0044] In this embodiment, by parsing the communication data of multiple group members sharing network access services of the target group, a feature data set of the target group is obtained. Based on the active time period and product interest set corresponding to each group member, the target group's product interest set corresponding to the current time is calculated. Based on the target group's product interest set, products that meet the interest threshold are selected as recommended products. Then, feature data of multiple shops are obtained. Based on the recommended products and the feature data of multiple shops, the comprehensive score of each shop is calculated. Shops whose comprehensive scores meet the score threshold are selected as target shops. Finally, based on the target group, recommended products, and target shops, an advertising generation model is applied to generate personalized advertisements and push them to the target group. Compared to existing technologies that only analyze the behavior of individual group members to recommend shops or products, this application deeply integrates the active time periods and product interests of all group members through shared network communication data of the target group. It generates a group-level product interest set that fits the current time period through quantitative calculation. When matching shops, it takes the overall needs of the group as the core, combines the positive and negative feature data of the shops to calculate the comprehensive score of the shops and select target shops. Finally, the personalized advertisements generated based on the target group, recommended products and target shops are more in line with the common needs and scenario characteristics of the group. After being pushed to the target group, it can achieve the dissemination effect of one person paying attention and the whole family being reached, thereby significantly improving the reach and exposure rate of shop advertisements.

[0045] Figure 2 A flowchart illustrating a group-oriented advertising push method according to another embodiment of this application is shown. Figure 2 As shown above, in the above Figure 1 Based on the illustrated embodiment, one implementation of step S102 is as follows: S201: Based on the active time period corresponding to each group member, select the first feature data set corresponding to the current time. The first feature data set includes the first group member feature data of multiple group members.

[0046] In this embodiment of the application, based on the active time period corresponding to each group member, a first feature data set corresponding to the current time is selected. The first feature data set includes first group member feature data of multiple group members. In one example, using... Representing the target group, assuming there are Each group member, then , Let the set of product interests contain the following products: For each product, .

[0047] S202: Based on the set of product interests in the first group member feature data corresponding to each group member, calculate the interest of each group member in each product and the purchase intention of each group member.

[0048] In this embodiment, the target group's interest in a particular product is jointly determined by the interest and purchase intention of each member within the group. A member's interest in a product represents the intensity of their attention to that product; while purchase intention reflects the degree to which that member is inclined to translate their interest into actual consumption behavior. Therefore, by calculating each member's interest in each product and their purchase intention separately, the target group's preference intensity and consumption probability for products can be quantified more comprehensively and accurately.

[0049] In one example, when product interest includes purchase frequency and interest behavior data, one specific implementation of step S202 is as follows: Based on interest behavior data, calculate the level of interest each group member has in each product.

[0050] In one example of this application embodiment, when the interest behavior data consists of the number of times a product is searched and added to cart or favorited, the interest level of each group member for each product can be calculated based on the number of searches and additions to cart or favorites for each product. Assume that the group members... For goods The number of keyword searches is The number of times you add items to your cart or favorites is [number]. The weighted method is used to calculate the group members' interest in the product. The calculation formula is as follows: (1) in, The weight corresponding to the number of searches. The weighting is assigned to the number of times items are added to cart or favorited, and... .

[0051] The set of each group member's interest in each product is then: (2) in, INST This represents the set of each group member's level of interest in each product; This represents the degree of interest of group member i in product j.

[0052] Based on purchase frequency and interest behavior data, the purchase intention of each group member is calculated.

[0053] In this embodiment of the application, the purchase intention of each group member is calculated based on purchase frequency and interest behavior data.

[0054] In this embodiment, interest behavior data characterizes the recent and dynamic interest preferences of group members, reflecting the changing trends of their current focus. Therefore, the interest level of each group member in each product calculated based on interest behavior data can directly reflect the intensity of the group member's current attention to the product. On the other hand, the number of purchases represents the stable behavioral pattern formed by the group members based on historical consumption, reflecting their long-term consumption tendencies and decision-making habits. Therefore, calculating the purchase intention of each group member based on the number of purchases and interest behavior data can accurately assess the group member's purchase intention, avoiding the problems of exaggerating demand by only looking at interest behavior or missing potential intentions by only looking at purchase records.

[0055] In one example, the specific calculation method for the purchasing intention of each group member is as follows: Based on the product interest set corresponding to each group member, obtain the total number of purchases of all products in the product interest set within a preset time window, as well as the total number of occurrences of interest behavior data for all products in the product interest set.

[0056] In this embodiment of the application, the total number of purchases of all products within the product interest set within a preset time window is obtained: (3) Where A represents the total number of purchases of all products within the product interest set within the preset time window. This represents the number of times group member i purchased product j; j represents the product identifier serial number; and m represents the quantity of the product.

[0057] Get the total number of occurrences of interest behavior data for all products in the product interest set within a preset time window: (4) Where B represents the total number of occurrences of interest behavior data for all products within the product interest set within the preset time window. This represents the number of times member i of the group searched for product j. This indicates the number of times group member i has added or favorited product j, where j represents the product's serial number; and m represents the quantity of the product.

[0058] In one example, the regular time corresponds to a 7-day time window, while the promotional time corresponds to a 3-day time window.

[0059] The ratio of the total number of purchases to the total number of occurrences of interest behavior data is calculated as the purchase intention of each group member.

[0060] In this embodiment of the application, the purchasing intention of each group member is assessed by calculating the ratio of the total number of purchases to the total number of occurrences of interest behavior data. In one example, the calculation formula is as follows: (5) in, A represents the purchase intention of group member i; i represents the group member identifier sequence number; A represents the total number of purchases of all products in the product interest set within the preset time window; B represents the total number of occurrences of interest behavior data for all products in the product interest set within the preset time window.

[0061] To eliminate interference from differences in behavioral frequency among members of different groups, purchasing intentions are uniformly mapped to a standardized numerical range, and the calculation formula is as follows: (6) in, This represents the normalized purchase intention of group member i; This indicates the purchasing intention of group member i; This represents the minimum purchasing intention of the group members; This represents the maximum purchasing intention of the group's members.

[0062] The normalized purchase intention matrix of the group members is then: ,in, This represents the normalized purchase intention matrix of the group members.

[0063] In this embodiment, based on the product interest set corresponding to each group member, the total number of purchases of all products in the product interest set within a preset time window, as well as the total number of occurrences of interest behavior data for all products in the product interest set, are obtained. This avoids interference from outdated consumption and behavior records on current purchase intentions, ensuring that the calculated purchase intentions are consistent with the current consumption status of group members. Then, the ratio of the total number of purchases to the total number of occurrences of interest behavior data is calculated as the purchase intention of each group member. This transforms the abstract total number of purchases and interest behavior data into quantifiable purchase intentions, thereby solving the problem of comparability of purchase intentions among members with different behavior frequencies.

[0064] S203: Based on each group member's interest in each product and each group member's willingness to purchase, calculate the target group's interest in each product, and construct a target group product interest set based on the target group's interest in each product.

[0065] In this embodiment of the application, the target group's interest in each product can be calculated based on each group member's interest in each product and each group member's purchase intention. In one example, the target group's interest in each product can be calculated based on each group member's interest in each product and the normalized purchase intention matrix of the group members. (7) in, This indicates the target group's level of interest in product j; This indicates the degree of interest of group member i in product j. This represents the normalized purchase intention of group member i; i represents the group member identifier sequence number; j represents the product identifier sequence number; and n represents the number of group members.

[0066] In one example, based on the characteristic data set of family groups shown in Table 1, the target group's product interest set corresponding to 6:00 to 9:00 is calculated (the calculation process takes...). As shown in Table 2, Table 2 is the target collective product interest set corresponding to the current time.

[0067]

[0068] In one example, at different time periods, the target collective product interest set includes: product identifiers and the corresponding interest levels for each product identifier. In another example, the target collective product interest set may also include the collective member ID corresponding to each product identifier. For example: During the time period from 6:00 to 7:00, the collective interest set of the family was: (baby and toddler products, 25.83025601, [1,2]), (mobile phone, 5.129557797, [1]), (snacks, 4.770209465, [2]), (sweater, 3.842668735, [2]), (lipstick, 3.172226532, [1]); During the time period from 7:00 to 8:00, the family's collective interest set of goods was: (baby and toddler products, 25.83025601, [1,2]) (mobile phone, 5.129557797, [1]), (snacks, 4.770209465, [2]), (sweater, 3.842668735, [2]), (lipstick, 3.172226532, [1]).

[0069] In this embodiment, based on the active time period corresponding to each group member, a first feature data set corresponding to the current time is selected, effectively eliminating data interference from inactive members and ensuring that the data used for calculation is highly compatible with the actual active status of the target group. Then, based on the product interest set in the first group member feature data corresponding to each group member, the interest level of each group member in each product and the purchase intention of each group member are calculated. Finally, based on the interest level of each group member in each product and the purchase intention of each group member, the interest level of the target group in each product is calculated, and a product interest set of the target group is constructed based on the interest level of the target group in each product. This accurately captures the core common needs of the target group, thereby providing a precise data foundation for subsequent product recommendation selection, store matching, and personalized advertising generation.

[0070] Figure 3 A flowchart illustrating a group-oriented advertising push method according to another embodiment of this application is shown. Figure 3 As shown above, in the above Figure 1 Based on the illustrated embodiment, one implementation of step S104 is as follows: S301: Based on the recommended products and the product list corresponding to each store, select the stores to be recommended. The product list corresponding to the stores to be recommended contains the recommended products.

[0071] In this embodiment, shops to be recommended are selected based on recommended products and the product list corresponding to each shop. In one example, the real-time business status of the shops and the characteristics of the active group members at the current time can be combined to further refine the selection of shops to be recommended. For example, only shops that are currently open for business can be selected, and the final set of shops to be recommended can be determined based on the matching degree between the types of active group members (such as office workers, students, the elderly, etc.) and the target audience of the shops.

[0072] S302: For each store to be recommended, the positive and negative feature data are standardized to obtain dimensionless positive and negative feature data; In this embodiment, for each store to be recommended, the positive and negative feature data corresponding to that store are standardized. In one example, thresholds are set for both positive and negative feature data. If the negative feature data exceeds the threshold or the positive feature data is less than the threshold, the group members will directly abandon the store. For example, the product price may be too high, the delivery time too long, or the rating too low. Since different feature data have different dimensions, they need to be dedivided before calculating the score. This application uses a ratio method to dedivide the positive and negative feature data, for example, by dividing the positive and negative feature data by their respective thresholds.

[0073] S303: Select the first function to normalize the dimensionless positive feature data to obtain normalized positive feature data. The first function is configured to increase the normalized positive feature data as the dimensionless positive feature data increases.

[0074] In this embodiment, a first function is selected to normalize the dimensionless positive feature data, mapping the dimensionless positive feature data to [0, 1]. In one example, the first function is the sigmoid function: (8) in, This represents the normalized value corresponding to the positive feature data j in shop i; Control the steepness of the first function curve; Control the center position of the first function curve; This represents the dimensionless value of the positive feature data in shop i.

[0075] S304: Select the second function to normalize the dimensionless negative feature data to obtain normalized negative feature data. The second function is configured to reduce the normalized negative feature data as the dimensionless negative feature data increases.

[0076] In this embodiment, a second function is selected to normalize the dimensionless negative feature data, mapping the dimensionless negative feature data to [0, 1]. In one example, the second function is the sigmoid function: (9) in, This represents the normalized value corresponding to the negative feature data j in shop i; Control the steepness of the second function curve; Control the center position of the second function curve; This represents the dimensionless value of the negative feature data in shop i.

[0077] S305: For each store to be recommended, a comprehensive score is calculated based on normalized positive feature data, normalized negative feature data, and their respective weights.

[0078] In this embodiment of the application, for each shop to be recommended, a comprehensive score is calculated based on normalized positive feature data, normalized negative feature data, and their respective weights. In one example, the calculation formula is: (10) in, This represents the overall score corresponding to shop i; This represents the weight corresponding to feature data j, where the feature data includes positive feature data and negative feature data; This represents the normalized value corresponding to the positive feature data j in shop i; This represents the normalized value corresponding to the negative feature data j in shop i.

[0079] In this embodiment, based on the recommended products and the product list corresponding to each store, stores to be recommended are selected, and stores without suitable products are directly excluded, avoiding redundant calculations for invalid stores and saving computing power costs. Then, the positive and negative feature data are standardized to make the multi-source heterogeneous data comparable. Then, the dimensionless positive and negative feature data are normalized based on the first and second functions respectively to ensure that the better the positive feature data, the higher the score, and the better the negative feature data, the lower the score. Finally, for each store to be recommended, a comprehensive score is calculated based on the normalized positive feature data, the normalized negative feature data, and their respective weights, realizing the quantitative evaluation and ranking of the multi-dimensional characteristics of the stores.

[0080] Figure 4 A flowchart illustrating a group-oriented advertising push method according to another embodiment of this application is shown. Figure 4 As shown above, in the above Figure 1 Based on the illustrated embodiment, one implementation of step S105 is as follows: S401: Obtain a preset advertising information template, which contains structured fields for generating advertising copy.

[0081] In this application embodiment, a preset advertising information template is obtained. In one example, the preset advertising information template of this application is: "Create an ad copy / video script / image, etc. for {Shop Name}. Focus on {Product Name} and its {Product Parameters}. Suitable for {Shop Setting Occasion}. Target audience is {Target Customer Group}. Please write using {Shop Style Settings}." The store style can be set by the store itself. It can be the overall style of the store or a seasonal feature, such as the Spring Festival or Valentine's Day. It can increase the interest of group members based on the relationship between the products and the season.

[0082] S402: Input the target audience, recommended products, and characteristic data of the target store as variables into the preset advertising information template to form advertising information.

[0083] In this embodiment of the application, the target group, recommended products and characteristic data corresponding to the target store can be used as variables and input into the preset advertising information template based on the structured fields in the preset advertising information template to form advertising information.

[0084] S403: Input advertising information into the advertising generation model to generate personalized advertisements that match the target store and the target group. The advertising generation model is a generative artificial intelligence model.

[0085] In this embodiment, a pre-trained ad generation model is used to generate personalized ads that match the target store and the target group. In one example, the ad generation model is a generative artificial intelligence model.

[0086] S404: Push personalized advertisements to the display terminals corresponding to the target group.

[0087] In this embodiment of the application, the generated personalized advertisement is pushed to the display terminal corresponding to the target group. In one example, when the target group is a family group, the personalized advertisement can be pushed to the family's large screen, such as a TV.

[0088] In this embodiment, a preset advertising information template is obtained, and the characteristic data corresponding to the target group, recommended products, and target stores are used as variables and input into the preset advertising information template to form advertising information. The advertising information is then input into the advertising generation model to generate personalized advertisements that match the target stores and target groups. This can quickly generate natural, smooth, and highly targeted personalized advertising content and push the personalized advertisements to the display terminals corresponding to the target groups. This achieves customized advertising push through "one policy per household and one policy per time" and precise reach of advertisements to shared scenarios of the groups, significantly improving the exposure rate and group awareness of advertisements.

[0089] Figure 5 This is an overall architecture diagram of a group-oriented advertising push system, such as... Figure 5 As shown, the data layer mines upstream traffic data through DPI data mining to obtain group member data, which is then stored in the database. Shop data is provided by shop owners during registration in the shop system. The functional layer, based on the group member data and shop data, obtains the feature data set of the target group and the feature data of the shops, respectively. Through the target group-shop matching function, it recommends shops to members of the target group. Then, based on the recommendation results, it generates corresponding advertising plans and pushes them to the target group members. At the group member layer, multiple members of the target group receive and obtain the recommendation results on the large-screen terminal. The large screen allows for the display of shared interests among group members, fully leveraging the advantages of the large-screen platform as a dissemination platform to provide a more intuitive and vivid display effect for shop advertising.

[0090] Figure 6 This is a schematic diagram of the overall process of a group-oriented advertising push system, such as... Figure 6 As shown, the following explanation uses a family group as an example: S601: Family Profile: Under home broadband, the mobile phone numbers and online behavior of each family member can be obtained, thereby creating a profile of each individual member of the family. Based on the family member profiles, the time distribution of family interest sets is extracted, and the family purchase rate of a certain product is calculated to complete the broadband family profile.

[0091] S602: Shop Feature Extraction: Construct shop tags and extract shop features. Shops can set style prompts to guide the generation of advertising plans that match the shop's style.

[0092] S603: Household and Shop Matching: Based on household profiles and product characteristics, assess the household purchase rate of products, match products with the highest household purchase rate to group members, rank shops based on price, rating, location, etc., and recommend top-ranked shops to group members, thereby achieving matching between shops and households.

[0093] S604: Prompt word generation: Combine the description of the store tag and the family profile to formulate prompt words for generating advertisements.

[0094] S605: Ad generation model: Input the prompt words into the ad generation model, complete the ad generation through the ad generation model, push it to the family group, and display it on the family group's large screen.

[0095] Figure 7 This is a schematic diagram of a group-oriented advertising push device provided in this application. Figure 2 As shown, the collective advertising push device 700 provided in this application includes: The first acquisition module 701 is used to obtain a set of characteristic data of the target group by parsing the communication data of multiple group members sharing the network access service. The set of characteristic data includes the characteristic data of multiple group members. Each group member's characteristic data includes: group member identifier, active time period, and set of product interests. The first calculation module 702 is used to calculate the target collective product interest set corresponding to the current time based on the active time period and product interest set corresponding to each collective member, and select products that meet the interest threshold as recommended products based on the target collective product interest set. The target collective product interest set includes: product identifier and the interest degree corresponding to the product identifier. The second acquisition module 703 is used to acquire feature data of multiple shops. The feature data of each shop includes: a product list, positive feature data that influences the selection of shops by group members, and negative feature data. The second calculation module 704 is used to calculate the comprehensive score of each store based on the feature data of recommended products and multiple stores, and select stores whose comprehensive scores meet the score threshold as target stores. The push module 705 is used to generate personalized ads based on the target audience, recommended products, and target stores using an advertising generation model, and then push them to the target audience.

[0096] In one example, the first computing module 702 includes: The selection module is used to select a first feature data set corresponding to the current time based on the active time period of each group member. The first feature data set includes the first group member feature data of multiple group members. The first calculation submodule is used to calculate each group member's interest in each product and each group member's purchase intention based on the product interest set in the first group member feature data corresponding to each group member. The second calculation submodule is used to calculate the target group's interest in each product based on each group member's interest in each product and each group member's willingness to purchase, and to construct a target group product interest set based on the target group's interest in each product.

[0097] In one example, the first computing module 702 includes: The first calculation submodule is also used to calculate the interest level of each group member for each product based on interest behavior data; The first calculation submodule is also used to calculate the purchase intention of each group member based on the number of purchases and interest behavior data.

[0098] In one example, the first computing module 702 includes: The first acquisition submodule is used to acquire the total number of purchases of all products in the product interest set within a preset time window, and the total number of occurrences of interest behavior data for all products in the product interest set, based on the product interest set corresponding to each group member. The first calculation submodule is also used to calculate the ratio of the total number of purchases to the total number of occurrences of interest behavior data, as a measure of each group member's purchase intention.

[0099] In one example, the second computing module 704 includes: The second acquisition submodule is used to select shops to be recommended based on the recommended products and the product list corresponding to each shop. The product list corresponding to the shops to be recommended contains the recommended products. The first processing submodule is used to standardize the positive and negative feature data for each recommended store to obtain dimensionless positive and negative feature data. The second processing submodule is used to select the first function to normalize the dimensionless positive feature data to obtain normalized positive feature data. The first function is configured such that the normalized positive feature data increases as the dimensionless positive feature data increases. The second processing submodule is also used to select the second function to normalize the dimensionless negative feature data to obtain normalized negative feature data. The second function is configured such that the normalized negative feature data decreases as the dimensionless negative feature data increases. The third calculation submodule is used to calculate a comprehensive score for each store to be recommended, based on normalized positive feature data, normalized negative feature data, and their respective weights.

[0100] In one example, push module 705 includes: The third acquisition submodule is used to acquire a preset advertising information template, which contains structured fields for generating advertising copy. The third processing submodule is used to take the characteristic data of the target group, recommended products and target stores as variables, input the preset advertising information template, and form advertising information. The generation module is used to input advertising information into the advertising generation model to generate personalized advertisements that match the target store and the target group. The advertising generation model is a generative artificial intelligence model. The push module is used to push personalized advertisements to the display terminals corresponding to the target group.

[0101] Figure 8 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application.

[0102] An electronic device may include a processor 801 and a memory 802 storing computer program instructions.

[0103] Specifically, the processor 801 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0104] Memory 802 may include mass storage for data or instructions. For example, and not limitingly, memory 802 may include a hard disk drive (HDD), a floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. In one instance, memory 802 may include removable or non-removable (or fixed) media, or memory 802 may be a non-volatile solid-state memory.

[0105] In one instance, memory 802 may be read-only memory (ROM). In one instance, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM), or flash memory, or a combination of two or more of these.

[0106] Memory 802 may include read-only memory (ROM), random access memory (RAM), disk storage media device, optical storage media device, flash memory device, electrical, optical, or other physical / tangible memory storage device. Therefore, generally, memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to one aspect of this disclosure.

[0107] The processor 801 implements a group-oriented advertising push method in the above embodiment by reading and executing computer program instructions stored in the memory 802.

[0108] In one example, the electronic device may also include a communication interface 803 and a bus 804. For example, Figure 8 As shown, the processor 801, memory 802, and communication interface 803 are connected through bus 804 and complete communication with each other.

[0109] The communication interface 803 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0110] Bus 804 includes hardware, software, or both, that couples components of an online data traffic metering device together. For example, and not as a limitation, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 804 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, this application contemplates any suitable bus or interconnect.

[0111] Furthermore, in conjunction with the fracturing pump fault detection method in the above embodiments, this application embodiment can provide a computer storage medium for implementation. The computer storage medium stores computer program instructions; when these computer program instructions are executed by a processor, they implement any of the group-oriented advertising push methods in the above embodiments.

[0112] This application also provides a computer program product, including a computer program, which, when executed, implements any of the group-oriented advertising push methods described in the above embodiments.

[0113] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0114] The functional blocks shown in the above block diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, read-only memory (ROM), flash memory, erasable read-only memory (EROM), floppy disks, compact disc read-only memory (CD-ROM), optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0115] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0116] The aspects of this disclosure have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by special-purpose hardware performing the specified functions or actions, or can be implemented by a combination of special-purpose hardware and computer instructions.

[0117] The above are merely specific embodiments of this application. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, modules, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. It should be understood that the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application.

Claims

1. A method for pushing advertisements to a group, characterized in that, include: By analyzing the communication data of multiple group members sharing network access services of the target group, a set of characteristic data of the target group is obtained. The set of characteristic data includes the characteristic data of multiple group members. Each characteristic data of the group member includes: group member identifier, active time period and product interest set. Based on the active time period and product interest set corresponding to each group member, calculate the target group product interest set corresponding to the current time, and select products that meet the interest threshold as recommended products based on the target group product interest set. The target group product interest set includes: product identifier and the interest level corresponding to the product identifier. Acquire feature data for multiple shops. The feature data for each shop includes: product list, positive feature data that influences the selection of shops by group members, and negative feature data. Based on the feature data of the recommended products and the multiple shops, calculate the comprehensive score of each shop, and select the shops whose comprehensive scores meet the score threshold as target shops. Based on the target group, recommended products, and target stores, an advertising generation model is used to generate personalized advertisements, which are then pushed to the target group.

2. The method according to claim 1, characterized in that, The calculation of the target group's product interest set corresponding to the current time, based on the active time period and product interest set of each group member, includes: Based on the active time period corresponding to each group member, a first feature data set corresponding to the current time is selected. The first feature data set includes the first group member feature data of multiple group members. Based on the set of product interests in the feature data of the first group member corresponding to each group member, calculate the interest of each group member in each product and the purchase intention of each group member; Based on each group member's interest in each product and their purchase intention, the target group's interest in each product is calculated, and a target group product interest set is constructed based on the target group's interest in each product.

3. The method according to claim 2, characterized in that, The product interest set corresponding to each group member includes multiple product interests, and each product interest includes: purchase frequency and interest behavior data; The calculation of each group member's interest in each product and their purchase intention based on the product interest set in the first group member feature data corresponding to each group member includes: Based on the interest behavior data, the interest level of each group member for each product is calculated; Based on the number of purchases and the interest behavior data, the purchase intention of each group member is calculated.

4. The method according to claim 3, characterized in that, The calculation of the purchase intention of each group member based on the purchase frequency and the interest behavior data includes: Based on the product interest set corresponding to each group member, obtain the total number of purchases of all products in the product interest set within a preset time window, as well as the total number of occurrences of interest behavior data for all products in the product interest set; The ratio of the total number of purchases to the total number of occurrences of the interest behavior data is calculated as the purchase intention of each group member.

5. The method according to claim 1, characterized in that, The step of calculating a comprehensive score for each store based on the feature data of the recommended products and the multiple stores, and selecting stores whose comprehensive scores meet the score threshold as target stores, includes: Based on the recommended products and the product list corresponding to each store, select stores to be recommended, wherein the product list corresponding to the stores to be recommended contains the recommended products; For each recommended store, the positive feature data and the negative feature data are standardized to obtain dimensionless positive feature data and negative feature data; The positive feature data after eliminating dimensions is normalized by a first function to obtain normalized positive feature data. The first function is configured such that the normalized positive feature data increases as the positive feature data after eliminating dimensions increases. The negative feature data after eliminating dimensions is normalized by a second function to obtain normalized negative feature data. The second function is configured such that the normalized negative feature data decreases as the negative feature data after eliminating dimensions increases. For each shop to be recommended, a comprehensive score is calculated based on the normalized positive feature data, the normalized negative feature data, and their respective weights.

6. The method according to claim 1, characterized in that, The step of generating personalized ads based on the target group, recommended products, and target stores using an advertising generation model, and then pushing them to the target group, includes: Obtain a preset advertising information template, wherein the preset advertising information template contains structured fields for generating advertising copy; The target group, the recommended products, and the characteristic data corresponding to the target store are used as variables and input into the preset advertising information template to form advertising information. The advertising information is input into the advertising generation model to generate personalized advertisements that match the target store and the target group, wherein the advertising generation model is a generative artificial intelligence model; The personalized advertisement is pushed to the display terminal corresponding to the target group.

7. A collective advertising push device, characterized in that, The device includes: The first acquisition module is used to acquire a set of feature data of the target group by parsing the communication data of multiple group members sharing the network access service. The set of feature data includes the feature data of multiple group members, and each feature data of the group member includes: group member identifier, active time period and product interest set. The first calculation module is used to calculate the target collective product interest set corresponding to the current time based on the active time period and product interest set of each collective member, and select products that meet the interest threshold as recommended products based on the target collective product interest set. The target collective product interest set includes: product identifier and the interest degree corresponding to the product identifier. The second acquisition module is used to acquire feature data of multiple shops. The feature data of each shop includes: a product list, positive feature data that influences the selection of shops by group members, and negative feature data. The second calculation module is used to calculate the comprehensive score of each store based on the feature data of the recommended products and the multiple stores, and select the stores whose comprehensive scores meet the score threshold as target stores. The push module is used to generate personalized advertisements based on the target group, recommended products, and target stores using an advertising generation model, and then push them to the target group.

8. An electronic device, characterized in that, The device includes: a processor and a memory storing computer program instructions; the processor reads and executes the computer program instructions to implement a group-oriented advertising push method as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer storage medium stores computer program instructions, which, when executed by a processor, implement a group-oriented advertising push method as described in any one of claims 1-6.

10. A computer program product, characterized in that, When the instructions in the computer program product are executed by the processor of the electronic device, the electronic device performs a group-oriented advertising push method as described in any one of claims 1-6.

Images