Coupon pushing method and device, electronic equipment and storage medium

By predicting the verification time and validity period to match users, and using user information to calculate the degree of interest, pushing coupons to target users, the problem of poor coupon pushing in the existing technology is solved, and the usage rate and promotion of coupons are improved Effect.

CN118967219AInactive Publication Date: 2024-11-15深圳市灵智数字科技有限公司
View PDF 7 Cites 0 Cited by

Patent Information

Application Number
CN202410905804.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-05
Publication Date
2024-11-15
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

When the prior art pushes coupons to users, it is difficult to accurately match users' shopping preferences and time preferences, resulting in some users not receiving or using the pushed coupons, affecting the promotion effect.

Method used

The basic user is determined by predicting the verification time and the validity period of the coupon, and the matching degree is calculated based on user information and coupon information, so as to select interested target users to push coupons to them.

Benefits of technology

This reduces the situation where users do not receive or use coupons, improves the push and promotion effects of coupons, and ensures that coupons are used effectively during the validity period.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118967219A_ABST
    Figure CN118967219A_ABST
Patent Text Reader

Abstract

The invention is suitable for the technical field of coupon recommendation, and provides a coupon pushing method and device, electronic equipment and a storage medium, and the method comprises the steps: determining a basic user according to the predicted cancel-after-verification time and the validity period of a to-be-pushed coupon, the predicted cancel-after-verification time is used for indicating the time when the user cancel-after-verification of the to-be-pushed coupon, the basic users comprise users whose predicted cancel-after-verification time is matched with the validity period; according to the user information of the basic user and the coupon information of the to-be-pushed coupon, the matching degree of the basic user and the to-be-pushed coupon is determined, and the matching degree is used for reflecting the interest degree of the basic user in the to-be-pushed coupon; determining a target user from the basic users based on the matching degree; and pushing the coupon to be pushed to the target user. According to the invention, the coupon pushing effect can be improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application belongs to the technical field of coupon recommendation, and in particular, relates to a coupon push method, device, electronic device, and computer-readable storage medium. Background Art

[0002] In order to increase sales, online platforms usually plan various promotional activities to attract users to shop. Among them, pushing coupons to users is a common promotional activity.

[0003] Currently, when coupons are pushed to users, a large number of coupons are usually pushed randomly to the majority of users. However, due to personal shopping preferences and shopping time preferences, some users usually do not claim or use the pushed coupons within the validity period of the pushed coupons, making it difficult for the promotional activities to achieve the expected promotional effect. Summary of the invention

[0004] The embodiments of the present application provide a coupon push method, device, electronic device, and storage medium, which can reduce the situation where users do not receive or do not use pushed coupons after receiving them, and improve the coupon push effect.

[0005] In a first aspect, an embodiment of the present application provides a coupon push method, including:

[0006] Determine the basic user according to the predicted redemption time and the validity period of the coupon to be pushed, wherein the predicted redemption time is used to indicate the time when the user redeems the coupon to be pushed, and the basic user includes the user whose predicted redemption time matches the validity period;

[0007] Determine, according to the user information of the basic user and the coupon information of the coupon to be pushed, the degree of matching between the basic user and the coupon to be pushed, wherein the degree of matching is used to reflect the interest degree of the basic user in the coupon to be pushed;

[0008] Determine a target user from the basic users based on the matching degree;

[0009] Push the coupon to be pushed to the target user.

[0010] In a second aspect, an embodiment of the present application provides a coupon push device, including:

[0011] A basic user determination module, used to determine basic users according to a predicted redemption time and a validity period of the coupon to be pushed, wherein the predicted redemption time is used to indicate a time when a user redeems the coupon to be pushed, and the basic users include users whose predicted redemption time matches the validity period;

[0012] A matching degree determination module, used to determine the matching degree between the basic user and the coupon to be pushed according to the user information of the basic user and the coupon information of the coupon to be pushed, wherein the matching degree is used to reflect the interest degree of the basic user in the coupon to be pushed;

[0013] A target user determination module, configured to determine a target user from the basic users based on the matching degree;

[0014] A push module is used to push the coupon to be pushed to the target user.

[0015] In a third aspect, an embodiment of the present application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the coupon push method described in the first aspect when executing the computer program.

[0016] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the coupon push method described in the first aspect are implemented.

[0017] In a fifth aspect, an embodiment of the present application provides a computer program product. When the computer program product is run on an electronic device, the electronic device executes the coupon push method described in the first aspect above.

[0018] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0019] In the embodiment of the present application, the basic user is determined based on the predicted redemption time and the validity period of the coupon to be pushed. Since the predicted redemption time can reflect the time when the user redeems the coupon to be pushed, the determined basic user is the user whose predicted redemption time matches the validity period of the coupon to be pushed. When determining the target user of the coupon to be pushed later, the target user is screened from the basic users, that is, the coupon to be pushed is pushed to the user who will redeem the coupon to be pushed within the validity period, reducing the situation where the user does not receive or does not use the pushed coupon after receiving it, ensuring the push effect of the coupon, that is, ensuring the promotion effect. At the same time, when determining the target user from the basic users, after determining the basic user's interest in the coupon to be pushed and obtaining the matching degree, the target user is determined from the basic users based on the matching degree, fully considering the personal preferences of the basic users, avoiding pushing to users who are not interested in the coupon to be pushed, and improving the push effect of the coupon. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art are briefly introduced below.

[0021] Figure 1 It is a flowchart of a coupon push method provided by an embodiment of the present application;

[0022] Figure 2 It is a structural diagram of a coupon pushing device provided in an embodiment of the present application;

[0023] Figure 3 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0024] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0025] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0026] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0027] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0028] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the phrases "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. appearing in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways.

[0029] Embodiment 1:

[0030] Figure 1 A schematic diagram of a coupon push method provided by an embodiment of the present invention is shown, and the details are as follows:

[0031] Step S101, determining basic users according to the predicted redemption time and the validity period of the coupon to be pushed, the predicted redemption time is used to indicate the time when the user redeems the coupon to be pushed, and the basic users include users whose predicted redemption time matches the validity period.

[0032] The above-mentioned predicted redemption time is the predicted time for indicating the user to redeem the coupon to be pushed. It should be noted that the predicted redemption time can be expressed as a specific moment or time period. Optionally, the predicted redemption time can be obtained by predicting the time when the user redeems the coupon to be pushed in combination with the user's shopping preferences through a model or manual method.

[0033] The validity period of the coupon to be pushed is the effective usage period of the coupon to be pushed, and the validity periods of different coupons are usually different.

[0034] Since there are usually certain rules for the shopping time and frequency of each user, in order to improve the push effect of coupons, before pushing coupons, it is possible to first determine whether the predicted redemption time of each user is within the validity period of the coupon to be pushed based on the user's predicted redemption time and the validity period of the coupon to be pushed, and filter out users whose predicted redemption time is within the validity period based on the judgment result to obtain basic users. Among them, the basic user is the user whose predicted redemption time is within the validity period. It can be considered that the basic user has a higher probability of receiving the coupon to be pushed and using the coupon to be pushed within the validity period.

[0035] In the embodiment of the present application, users whose predicted redemption time matches the validity period of the coupon to be pushed are first screened out to obtain basic users, so that when the target users of the coupon to be pushed are subsequently screened, the basic users who have a higher probability of redeeming the coupon to be pushed within the validity period are screened, which can reduce the situation where users do not claim the pushed coupons or do not use them after claiming them, thereby ensuring the push effect of the coupons.

[0036] Step S102, determining the matching degree between the basic user and the coupon to be pushed according to the user information of the basic user and the coupon information of the coupon to be pushed, wherein the matching degree is used to reflect the interest degree of the basic user in the coupon to be pushed.

[0037] The user information at least includes the user's identity identifier (such as an ID number or user ID, etc.). Optionally, the user information may also include one or more of the following information: gender, age, location, average monthly consumption, consumption level, user behavior sequence, core word sequence, etc. The user behavior sequence may include the product identifier (such as product number or name, etc.) of the product corresponding to each user's behavior, and the core word sequence may include the core word of the product corresponding to the user behavior sequence.

[0038] The above coupon information includes at least a coupon identifier (such as a coupon number, etc.), and may also include one or more of the usage threshold, discount amount, inventory, push time and validity period of the coupon to be pushed.

[0039] Since different users have different shopping preferences and different levels of interest in coupons, in order to further improve the push effect of coupons, after determining the basic users, the user information of the basic users and the coupon information of the coupons to be pushed can be combined to analyze the degree of interest of the determined basic users in the coupons to be pushed, and the degree of match between the basic users and the coupons to be pushed can be obtained. This allows for subsequent recommendations based on the interest levels of different basic users in the coupons to be pushed when determining the target users for the coupons to be pushed, thereby improving the user experience.

[0040] In some embodiments, the degree of match between the basic user and the coupon to be pushed can be determined based on the user information, the coupon information of the coupon to be pushed, and the applicable product information, and the applicable product information at least includes the central word of the product to which the coupon to be pushed applies, so that when analyzing the degree of match between the basic user and the coupon to be pushed, the basic user's interest in the coupon to be pushed can be comprehensively analyzed from the coupon dimension and the applicable product dimension, thereby improving the comprehensiveness of the calculated degree of match. In the embodiment of the present application, the user information also includes a central word sequence, and the degree of match between the basic user and the coupon to be pushed is analyzed in the central word dimension in combination with the products that the basic user has paid attention to and the applicable products of the coupon to be pushed, thereby further improving the accuracy of the obtained degree of match.

[0041] Optionally, the above applicable product information also includes one or more of application scenario, category, brand, price and product identification information.

[0042] Step S103: determining a target user from the basic users based on the matching degree.

[0043] Optionally, basic users whose matching degree is greater than or equal to a matching degree threshold (such as 0.7) can be determined as target users, so that the determined target users are users with a higher degree of interest in the coupons to be pushed, so as to increase the redemption rate of the coupons to be pushed, and at the same time, enable the coupons that the user is interested in to be pushed to the user in the future, thereby improving the user experience.

[0044] In some embodiments, when the number of basic users whose matching degree is greater than or equal to the matching degree threshold is greater than the expected number of coupons to be pushed, the basic users whose matching degree is greater than or equal to the matching degree threshold can be screened so that the number of determined target users meets the expected number of pushes, thereby avoiding the situation where the number of target users exceeds the number of coupons that can be pushed.

[0045] Step S104: Push the coupon to be pushed to the target user.

[0046] Optionally, when pushing the coupons to be pushed to the target users, the coupons to be pushed can be pushed to the target users by one or more push methods such as sending messages to the target users within the sales platform, pushing links in the target users' background, displaying banners at the top of the sales platform, specific push pages, and sending text messages. No specific restrictions are made here.

[0047] In the embodiment of the present application, in the process of determining the target user of the coupon to be pushed, the basic user is determined according to the predicted verification time and the validity period of the coupon to be pushed. Since the predicted verification time can reflect the time when the user verifies the coupon to be pushed, the determined basic user is the user whose predicted verification time matches the validity period of the coupon to be pushed. When determining the target user of the coupon to be pushed, the target user is screened from the basic users, that is, the coupon to be pushed is pushed to the user who will verify the coupon to be pushed within the validity period, reducing the situation that the user does not receive or does not use the pushed coupon after receiving it, ensuring the push effect of the coupon, that is, ensuring the promotion effect. At the same time, when determining the target user from the basic users, after determining the interest level of the basic user in the coupon to be pushed and obtaining the matching level, the target user is determined from the basic users in combination with the matching level, fully considering the personal preferences of the basic users, avoiding pushing to users who are not interested in the coupon to be pushed, further reducing the situation that the user does not receive the pushed coupon or does not use the pushed coupon after receiving it, and improving the push effect of the coupon.

[0048] In some embodiments, the above step S102 includes:

[0049] The user information of the above-mentioned basic users and the above-mentioned coupon information are used as inputs of the pre-trained analysis model to obtain the above-mentioned matching degree output by the above-mentioned analysis model. The above-mentioned analysis model is used to analyze the interest degree of the above-mentioned basic users in the above-mentioned coupons to be pushed according to the above-mentioned user information and the above-mentioned coupon information to obtain the above-mentioned matching degree.

[0050] In the embodiment of the present application, the pre-trained analysis model is used to automatically analyze the interest of basic users in the coupons to be pushed, and the required matching degree is obtained, thereby improving the analysis efficiency. In addition, the analysis model can better mine the deep relationship between user information and coupon information, so that the obtained matching degree has better accuracy.

[0051] In some embodiments, the input of the analysis model also includes applicable product information, so that the analysis model can comprehensively analyze the basic users' interest in pushing coupons from the dimensions of coupons and applicable products, thereby improving the comprehensiveness of the obtained matching degree.

[0052] Optionally, the above analysis model can be an attention-based neural network model. By embedding attention, the analysis model pays more attention to important information in user information and coupon information during the analysis process, improves the understanding ability of the analysis model, and thus improves the accuracy of the analysis model.

[0053] In other embodiments, the analysis model may also be a large language model, and a large language model with a large number of parameters can better analyze the sentiment tendencies in user information and extract key information in user information and coupon information, better understand the relationship between user preferences and coupons, and thus better analyze the basic users' interest in pushing coupons and improve the accuracy of the obtained matching degree.

[0054] In other embodiments, the analysis model can also be a dual-tower model consisting of a user tower, a coupon tower and an interoperability layer. The user tower stores user feature vectors obtained by pre-extracting feature information of basic users, and the coupon tower stores coupon feature vectors obtained by extracting feature information of coupons. Each user feature vector and coupon feature vector is then used as input to the interoperability layer, and the required degree of interest is calculated through the interoperability layer. The dual-tower model is used to fully explore the deep relationship between the coupons to be pushed and the basic users. At the same time, there is no need to extract the user feature vectors corresponding to each basic user in real time, which effectively improves the computing efficiency.

[0055] In some embodiments, before the above step of using the user information of the basic user and the coupon information as inputs of the pre-trained analysis model, the method further includes:

[0056] A1. Obtain first user behavior data, where the first user behavior data includes user behavior data for pushing products and / or pushing coupons.

[0057] In an embodiment of the present application, the sales platform will push content to the user, and the pushed content may include goods and coupons. Here, the user's behavior data for the pushed content is obtained as the first user behavior data. Among them, when the user has paid attention to the pushed goods in the pushed content (such as the user viewing, collecting or purchasing the pushed goods), the first user behavior data includes the user's behavior data for the pushed goods; when the user has paid attention to the pushed coupons in the pushed content (such as the user viewing, receiving or using the pushed coupons), the first user behavior data includes the user's behavior data for the pushed coupons. It can be understood that when the user has paid attention to the pushed goods in the pushed content and the pushed coupons, the first user behavior data may include the user's behavior data for the pushed goods, and the user's behavior data for the pushed coupons.

[0058] Optionally, the behavior data of each user for pushing goods and / or pushing coupons within a first preset time period (such as the past month) may be obtained as the first user behavior data corresponding to each user.

[0059] In some embodiments, the first user behavior data may include the behavior data of the user's additional purchase and purchase behavior for the pushed goods, and / or, the first user behavior data may include the behavior data of the user's use behavior (also called redemption behavior) for the pushed coupons. That is, the behavior data of the user's first behavior for the pushed goods and / or pushed coupons (such as additional purchase behavior for goods and redemption behavior for coupons, etc.) are obtained, and these specific behaviors can better reflect the user's shopping preferences and preferences for coupons, thereby improving the quality of the training samples constructed according to the first user behavior data, and further, can help the analysis model learn more accurate features and rules, and improve the training effect and accuracy of the analysis model.

[0060] In some embodiments, when the sales platform pushes information to the user, the pushed products do not belong to the products to which the pushed coupons in the pushed content are applicable, so as to increase the diversity of the products corresponding to the pushed content, and at the same time reduce the duplication of the candidate coupons for the pushed products determined subsequently and the pushed coupons in the first user behavior data, thereby improving the construction effect of the coupon samples.

[0061] A21. When the first user behavior data includes the pushed coupon, construct a coupon sample based on the pushed coupon.

[0062] In an embodiment of the present application, when the first user behavior data includes the user's behavior data for pushed coupons, it indicates that the user is more interested in the pushed coupons. At this time, a coupon sample can be constructed based on the pushed coupons corresponding to the user's actual behavior (i.e., the pushed coupons in the first user behavior data). Subsequently, the analysis model can be trained using the coupon samples constructed based on the user's actual behavior for pushed coupons to ensure the accuracy of the trained analysis model.

[0063] In some embodiments, when constructing a coupon sample based on a pushed coupon, the coupon sample is constructed based on the coupon information of the pushed coupon. Optionally, the label of the coupon sample can be determined based on the receipt status or cancellation status of the pushed coupon. For example, when focusing on the cancellation rate of the pushed coupon, the coupon sample corresponding to the pushed coupon cancelled by the user can be used as a positive sample (such as the sample label is 1), and the coupon sample corresponding to the pushed coupon not cancelled by the user can be used as a negative sample (such as the sample label is 0). For another example, when focusing on the promotion effect of the pushed coupon, the coupon sample corresponding to the pushed coupon received by the user can be used as a positive sample, and the coupon sample corresponding to the pushed coupon not received by the user can be used as a negative sample. The specific setting can be made according to actual needs.

[0064] A22. In the case where the first user behavior data includes the pushed product, construct the coupon sample according to the candidate coupons corresponding to the pushed product, and the similarity between the candidate coupons and the pushed product is greater than or equal to a similarity threshold.

[0065] In the embodiment of the present application, since different users have different preferences for products and coupons, their attention to pushed products and coupons varies. Some users may pay more attention to pushed products and less attention to pushed coupons, resulting in less real behavior data of users for pushed coupons. Therefore, in order to improve the training effect of the analysis model, when the first user behavior data includes pushed products, the candidate coupons corresponding to the pushed products can be determined first, and then a coupon sample can be constructed based on the candidate coupons.

[0066] Since pushed products can reflect users' shopping preferences, and candidate coupons are coupons whose similarity with pushed products is greater than or equal to a similarity threshold (the first similarity threshold, such as 0.9), they meet users' shopping preferences to a certain extent. Therefore, constructing coupon samples based on candidate coupons corresponding to pushed products can solve the problem of a small number of coupon samples due to little or no user behavior data for pushed coupons, while ensuring the quality of the constructed coupon samples.

[0067] In some embodiments, when constructing a coupon sample based on candidate coupons corresponding to pushed products, the coupon sample can be constructed based on the coupon information of the candidate coupons, and then the label of the constructed coupon sample is determined based on the purchase status of the pushed products; alternatively, the label of the coupon sample can also be determined based on the collection status or redemption status of the corresponding candidate coupons.

[0068] A3. The constructed analysis model is trained based on the constructed coupon samples and the corresponding user information to obtain the pre-trained analysis model.

[0069] After increasing the number of coupon samples through the above steps, the constructed analysis model can be trained according to the constructed coupon samples and the user information corresponding to the coupon samples, so that the constructed analysis model can fully learn the deep relationship between the pushed coupons in the coupon samples and the corresponding user information, and obtain a pre-trained analysis model with good accuracy.

[0070] Optionally, when the number of iterations of the analysis model reaches a preset number (such as 100 times) and / or the accuracy of the analysis model reaches an accuracy threshold (such as 0.95), the training of the analysis model can be stopped to obtain a pre-trained analysis model to ensure the accuracy of the obtained analysis model.

[0071] In some embodiments, before the above step A22, the method further includes:

[0072] B1. Acquire second user behavior data corresponding to each user, where the second user behavior data includes historical commodities and historical coupons corresponding to each behavior of the user.

[0073] In the embodiment of the present application, the second user behavior data is based on the user's behavior data on the products and coupons of the entire sales platform.

[0074] In other embodiments, the historical coupons in the second user behavior data may be coupons other than the pushed coupons, so as to reduce the probability that the candidate coupons subsequently determined based on the historical coupons are repeated with the pushed coupons in the first user behavior data, thereby improving the efficiency of constructing coupon samples.

[0075] Optionally, the second user behavior data may include historical products corresponding to the user's add-to-cart behavior and purchase behavior, and historical coupons corresponding to the coupon redemption behavior, that is, the second user behavior data can be determined based on the user's second behavior (such as product add-to-cart behavior and purchase behavior, and coupon collection behavior or redemption behavior).

[0076] B2. For each of the historical commodities, determine the candidate coupons corresponding to the historical commodities based on the historical coupons, wherein the candidate coupons include the historical coupons whose similarity with the historical commodities is greater than or equal to the similarity threshold.

[0077] B3. The historical products and the corresponding candidate coupons are associated and stored.

[0078] Correspondingly, the above-mentioned coupon sample is constructed according to the candidate coupons corresponding to the above-mentioned pushed products, including:

[0079] The target historical product is determined according to the historical products that match the pushed product, and the coupon sample is constructed according to the candidate coupons associated with the target historical product.

[0080] After determining multiple historical products and multiple historical coupons based on the second user behavior data of each user, for each historical product, calculate the similarity between the historical product and each historical coupon, and then screen out historical coupons whose similarity to the historical product is greater than or equal to a similarity threshold (such as 0.9), and use them as candidate coupons corresponding to the historical product. The determined candidate coupons are then associated with the historical product and stored to facilitate subsequent search for candidate coupons corresponding to the historical product.

[0081] In some embodiments, when there are a large number of historical coupons whose similarity to the historical product is greater than or equal to the similarity threshold, these historical coupons can be screened and a preset number (such as 3) of historical coupons can be retained as candidate coupons corresponding to the historical product.

[0082] For example, the selection probability corresponding to each historical coupon can be calculated, and the three historical coupons with higher selection probability can be used as candidate coupons corresponding to the historical products. Alternatively, the three historical coupons with higher similarity among the historical coupons whose similarity to the historical products is greater than or equal to the similarity threshold can be used as candidate coupons.

[0083] Optionally, for each historical product, a candidate coupon corresponding to the historical product can be determined based on the historical coupon corresponding to the second user behavior data corresponding to the historical product, so as to improve the correlation between the candidate coupon and the historical product, that is, to improve the correlation between the coupon sample subsequently constructed based on the candidate coupon and the user; or, the candidate coupon corresponding to the historical product can be determined based on all historical coupons, so as to increase the number of historical products with corresponding candidate coupons, and then increase the number of coupon samples subsequently constructed.

[0084] It can be understood that the pushed product is determined from various products on the sales platform. The user may have paid attention to the product when it was not a pushed product, that is, the pushed product and the historical product may be the same product. Therefore, when determining the candidate coupons associated with the pushed product, you can first determine whether there are historical products that match the pushed product. If there are historical products that match the pushed product, you can use this historical product that matches the pushed product as the target historical product, and then obtain the candidate coupons stored in association with the target historical product to construct a coupon sample, thereby ensuring the similarity between the candidate coupons used to construct the coupon sample and the pushed product.

[0085] Optionally, when determining the target historical product based on the historical products that match the pushed product, the historical product with the same product identifier as the pushed product can be used as the target historical product; or, when there is no historical product with the same product identifier as the pushed product, the historical product with the greatest similarity to the pushed product and the similarity greater than or equal to a second similarity threshold (such as 0.98) can be used as the target historical product. On the basis of ensuring the quality of the constructed coupon samples, the number of constructed coupon samples can be increased as much as possible.

[0086] In some embodiments, the above steps of constructing the above coupon sample according to the above candidate coupons associated with the target historical product include:

[0087] C1. When the number of the candidate coupons associated with the target historical product is greater than 1, determine the target candidate coupon based on the selection probability corresponding to the candidate coupons, where the selection probability reflects the probability of the candidate coupon being selected by the user.

[0088] C2. Construct the above coupon sample based on the above target candidate coupon.

[0089] To improve the quality of the constructed coupon samples, if the number of candidate coupons associated with the target historical product is greater than 1, we can first analyze the probability of each candidate coupon being selected by the user corresponding to the pushed product to obtain the selection probability of the candidate coupons. Then, we can combine the selection probabilities of the candidate coupons to screen out candidate coupons with a higher probability of being selected by the user from among the candidate coupons to obtain the target candidate coupons.

[0090] When constructing coupon samples, the coupon samples may be constructed based on target candidate coupons that are more likely to be selected by the user, so that the target candidate coupons in the constructed coupon samples are closer to the actual preferences of the user.

[0091] In an embodiment of the present application, when the number of candidate coupons associated with the target historical product is greater than 1, the probability of the candidate coupons being selected by the user corresponding to the pushed product is analyzed, and then the target candidate coupons for constructing the coupon sample are screened out based on the selection probability, thereby improving the matching degree between the coupons in the constructed coupon sample and the user, thereby improving the quality of the constructed coupon sample.

[0092] In some embodiments, before determining the target candidate coupon based on the selection probability corresponding to the candidate coupon, the method further includes:

[0093] D1. For each candidate coupon associated with the target historical product, a first score and a second score are determined based on the pushed product and the candidate coupon, wherein the first score reflects the degree of price matching between the pushed product and the candidate coupon, and the second score reflects the degree of matching between the consumption level of the user corresponding to the pushed product and the candidate coupon.

[0094] Optionally, the first score of the candidate coupon can be determined based on the price of the pushed product and the price corresponding to the usage threshold of the candidate coupon; alternatively, the price difference between the price of the pushed product and the usage threshold of the candidate coupon can be calculated (the price difference is an absolute value), and the first score of the candidate coupon can be determined based on the ratio of the price difference to the usage threshold.

[0095] For example, assume that the calculation rules of the first score are as follows: when the ratio of the price difference to the usage threshold is less than or equal to the first threshold (such as 0.2), the first score is 100; when the ratio of the price difference to the usage threshold is greater than the first threshold and less than or equal to the second threshold (such as 1), the first score is 80; when the ratio of the price difference to the usage threshold is greater than the second threshold and less than or equal to the third threshold (such as 5), the first score is 60; when the ratio of the price difference to the usage threshold is greater than the third threshold, the first score is 0.

[0096] Optionally, the second score of the candidate coupon can be determined based on the difference between the average consumption amount of the user corresponding to the pushed product in the second preset time period (such as the average consumption amount in the past month) and the use threshold of the candidate coupon. The first preset time period and the second preset time period can be the same time period or different time periods, which is not specifically limited here.

[0097] D2. Determine the selection probability corresponding to the candidate coupon according to the first score, the second score, the third score, the fourth score and the weights corresponding to the scores, wherein the third score reflects the popularity of the candidate coupon, and the fourth score reflects the similarity between the candidate coupon and the target historical product.

[0098] The weights corresponding to the first score, the second score, the third score and the fourth score may be pre-set fixed weights or dynamic weights that are dynamically adjusted according to actual application requirements.

[0099] Optionally, the third score of the candidate coupon may be calculated based on the total number of times the candidate coupon is clicked or used and the total number of times the candidate coupon is exposed.

[0100] In some embodiments, when determining the selection probability corresponding to a candidate coupon, the target weight corresponding to the candidate coupon can be calculated based on the sensitivity (first sensitivity) of the user corresponding to the candidate coupon (first user) to the candidate coupon, the sensitivity of the historical user to the coupon (second sensitivity) and the sensitivity of the user corresponding to the pushed product (second user) to the coupon (third sensitivity).

[0101] Among them, the first user's sensitivity to the candidate coupon can be determined based on the ratio of the number of times the first user clicks on the candidate coupon and the number of times the candidate coupon is exposed to the first user. The first user's sensitivity to the coupon can be determined based on the ratio of the total number of times the first user clicks or uses the coupon and the total number of times each coupon clicked or used by the first user is exposed to the first user. Similarly, the second user's sensitivity to the coupon is calculated according to the same steps, which will not be repeated here.

[0102] In the process of determining the selection probability, the total score corresponding to the candidate coupon can be first calculated based on the first score, the second score, the third score, the fourth score and the weights corresponding to the scores, and then the selection probability corresponding to the candidate coupon can be calculated based on the total score and the target weight, fully considering the user's sensitivity to the coupon and the attractiveness of the candidate coupon to the user, so as to improve the accuracy of the obtained rotation probability.

[0103] In the embodiment of the present application, a comprehensive analysis is performed based on multiple indicators such as the price matching degree between the pushed product and the candidate coupons and the popularity of the candidate coupons to obtain the probability that the candidate coupon will be selected by the user corresponding to the pushed product, thereby improving the accuracy of the calculated selection probability, that is, improving the accuracy of the target candidate coupons, thereby improving the accuracy of the constructed coupon samples.

[0104] In some embodiments, before the above step S101, the method further includes:

[0105] The user order data is used as the input of the pre-trained time prediction model to obtain the predicted redemption time output by the time prediction model. The user order data is used to reflect the user's commodity purchase status and coupon redemption status. The time prediction model is used to predict the time when the user will redeem the coupon to be pushed based on the user order data to obtain the predicted redemption time.

[0106] The above-mentioned user order data at least includes the goods purchased by the user (purchased goods) and the purchase time, as well as the redeemed coupons corresponding to the purchased goods. The redemption time of the coupons is the purchase time of the corresponding purchased goods.

[0107] Optionally, the input of the above time prediction model also includes user information, so that the time prediction model can predict the time when the user redeems the pushed coupon in combination with the characteristics of the user dimension, further improving the accuracy of the predicted redemption time.

[0108] Optionally, the time prediction model also includes order-related information, which includes the location associated with the time (such as the location when the goods were purchased), weather (such as the weather when the goods were purchased), day of the week, season and festival, etc., so that the time prediction model can combine the associated information when the user purchases the goods during the prediction process to fully consider the impact of the environment and specific time on the user's shopping time, thereby improving the accuracy of the predicted verification time.

[0109] Optionally, the order data of the user in a third preset time period (such as the past three months) may be obtained as the user order data.

[0110] Optionally, the time prediction model can be a gradient boosting model (Light Gradient Boosting Machine, LightGBM), and the LightGBM model uses a decision tree as a base learner, which can improve the generalization ability of the time prediction model in time series prediction by regularization means, and better learn the features of time-related user order data, thereby improving the accuracy of the predicted cancellation time. The time prediction model can also be an autoregressive model or a long short-term memory network (Long Short-Term Memory, LSTM), etc., which can be set according to the actual scenario requirements.

[0111] In the embodiment of the present application, based on the user's historical order data, the time for the user to redeem the coupon to be pushed is automatically analyzed through a time prediction model, which can improve the efficiency and accuracy of the predicted redemption time.

[0112] In some embodiments, the time prediction model is also used to predict the time when the user will receive the coupon to be pushed, and obtain the predicted time of receipt. The step S104 includes:

[0113] The coupon to be pushed is pushed to the target user based on the corresponding predicted collection time.

[0114] Since different users usually have different preferences for online time and coupon collection time, in order to improve the collection rate of pushed coupons, before pushing the coupons to be pushed to the target users, the time when the target users will collect the coupons to be pushed can be predicted through a time prediction model to obtain the predicted collection time.

[0115] When pushing the coupon to be pushed to the target user, the coupon to be pushed is pushed to the target user according to the corresponding predicted collection time of the target user, that is, the coupon to be pushed is pushed to the target user at a time when the probability of the target user collecting the coupon to be pushed is relatively high, thereby realizing personalized push of the coupon to be pushed, increasing the probability of the coupon to be pushed being collected, and helping to improve the redemption rate of the coupon to be pushed.

[0116] In some embodiments, before the above step S101, the method further includes:

[0117] Determine from each user a user whose user tag matches the delivery tag corresponding to the coupon to be pushed, and obtain the screened user, wherein the user tag includes at least one of a consumption level tag and an activity level tag.

[0118] Correspondingly, the above step S101 includes:

[0119] According to the predicted redemption time and the above validity period, the above basic users are determined from the above screened users.

[0120] The user tags include at least one of a consumption level tag and an activity level tag, so as to filter out users whose consumption levels match the usage threshold of the coupons to be pushed, or to filter out users with high or low recent activity.

[0121] Optionally, when determining the consumption level label of the user, the consumption data of the user may be processed by clustering analysis (such as k-means clustering algorithm) to obtain the consumption level label of the user.

[0122] Optionally, when determining the activity level label, the user's activity data may be processed by cluster analysis to obtain the user's activity level label; or, the user's activity level label may be determined based on the number of occurrences of different user behaviors.

[0123] For example, the number of times the user clicks, adds to cart, and places an order in a fourth preset time period (e.g., the past three months) can be counted, and then the user's activity level label is calculated based on the number of times each behavior occurs and the corresponding weight. Among them, the weight of the ordering behavior is greater than the weight corresponding to the add to cart behavior, and the weight corresponding to the add to cart behavior is greater than the weight corresponding to the click behavior.

[0124] Optionally, the user identification may also include a city tag (such as a city that the user has visited within the fourth preset time period), a gender tag, an age tag,

[0125] In order to further improve the accuracy of the target users finally determined, in an embodiment of the present application, one or more user tags corresponding to each user are set. When determining the coupon to be pushed, the delivery tag corresponding to the coupon to be pushed can be set, and then the coupon to be pushed can be pushed to the user whose user tag matches the delivery tag.

[0126] For example, in order to improve the retention rate of users of the sales platform, the user tag may include an activity tag, and the values ​​of the activity tag include the following three types: high activity, medium activity and low activity; when setting the delivery tag of coupon A to be pushed, the delivery tag is set to: users with low activity. Then, when screening each user, based on the delivery tag, the values ​​of the activity tag are screened out as: users with low activity as the screened users, so that the final target user is the low-activity user, and coupon A to be pushed is pushed to the low-activity user to re-stimulate the interest of the low-activity user and improve user activity.

[0127] Among them, before determining the basic users, users whose user tags match the delivery tags of the coupons to be pushed can be screened out from various users to obtain the screened users, so that when the basic users are subsequently determined, the basic users are directly determined from the screened users, that is, it is only necessary to predict the predicted redemption time of the screened users, and then compare the predicted redemption time of the screened users with the validity period of the coupons to be pushed. There is no need to predict the predicted redemption time for each user, which effectively reduces the processing volume and improves the efficiency of determining target users.

[0128] Optionally, the delivery tags of the coupons to be pushed may include multiple ones. When filtering out users whose user tags match the delivery tags from various users based on the delivery tags of the coupons to be pushed, users whose user tags match the delivery tags are filtered out from various users based on the filtering logic (such as union) corresponding to each delivery tag and the coupons to be pushed.

[0129] For example, assuming that the delivery tags include delivery tag A and delivery tag B, and the filtering logic is the difference set filtering of delivery tag A and delivery tag B, then when filtering users, users whose user tags match delivery tag A but do not match delivery tag B, and users whose user tags do not match delivery tag A but match delivery tag B are filtered out and regarded as the filtered users.

[0130] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0131] Embodiment 2:

[0132] Corresponding to the coupon push method described in the above embodiment, Figure 2 A structural block diagram of a coupon push device provided in an embodiment of the present application is shown. For ease of explanation, only the parts related to the embodiment of the present application are shown.

[0133] Reference Figure 2 The device includes: a basic user determination module 21, a matching degree determination module 22, a target user determination module 23 and a push module 24.

[0134] The basic user determination module 21 is used to determine the basic users according to the predicted redemption time and the validity period of the coupon to be pushed. The above-mentioned predicted redemption time is used to indicate the time when the user redeems the above-mentioned coupon to be pushed. The above-mentioned basic users include users whose above-mentioned predicted redemption time matches the above-mentioned validity period.

[0135] The matching degree determination module 22 is used to determine the matching degree between the basic user and the coupon to be pushed according to the user information of the basic user and the coupon information of the coupon to be pushed, wherein the matching degree is used to reflect the interest degree of the basic user in the coupon to be pushed.

[0136] The target user determination module 23 is used to determine the target user from the basic users based on the matching degree.

[0137] The push module 24 is used to push the coupons to be pushed to the target users.

[0138] In the embodiment of the present application, in the process of determining the target user of the coupon to be pushed, the basic user is first determined based on the predicted verification time and the validity period of the coupon to be pushed, and the predicted verification time can reflect the time when the user verifies the coupon to be pushed. The determined basic user is the user whose predicted verification time matches the validity period of the coupon to be pushed. When determining the target user of the coupon to be pushed, the target user is screened from the basic users, that is, the coupon to be pushed is pushed to the user who will verify the coupon to be pushed within the validity period, reducing the situation that the user does not receive or does not use the pushed coupon after receiving it, ensuring the push effect of the coupon, that is, ensuring the promotion effect. At the same time, when determining the target user from the basic users, after determining the basic user's interest in the coupon to be pushed and obtaining the matching degree, the target user is determined from the basic users in combination with the matching degree, fully considering the basic user's personal preferences, avoiding pushing to users who are not interested in the coupon to be pushed, further reducing the situation that the user does not receive the pushed coupon or does not use the pushed coupon after receiving it, and improving the push effect of the coupon.

[0139] In some embodiments, the matching degree determination module 22 includes:

[0140] The model processing unit is used to use the user information of the above-mentioned basic users and the above-mentioned coupon information as inputs of the pre-trained analysis model to obtain the above-mentioned matching degree output by the above-mentioned analysis model. The above-mentioned analysis model is used to analyze the interest degree of the above-mentioned basic users in the above-mentioned coupons to be pushed based on the above-mentioned user information and the above-mentioned coupon information to obtain the above-mentioned matching degree.

[0141] In some embodiments, the coupon push device further includes:

[0142] The first user behavior data acquisition module is used to acquire first user behavior data, where the first user behavior data includes user behavior data for pushing products and / or pushing coupons.

[0143] The first constructing module is used to construct a coupon sample based on the pushed coupon when the first user behavior data includes the pushed coupon.

[0144] The second construction module is used to construct the above-mentioned coupon sample according to the candidate coupons corresponding to the above-mentioned pushed products when the above-mentioned first user behavior data includes the above-mentioned pushed products, and the similarity between the above-mentioned candidate coupons and the above-mentioned pushed products is greater than or equal to the similarity threshold.

[0145] The training module is used to train the constructed analysis model based on the constructed coupon samples and the corresponding user information to obtain the pre-trained analysis model.

[0146] In some embodiments, the coupon push device further includes:

[0147] The second user behavior data acquisition module is used to acquire the second user behavior data corresponding to each user, and the second user behavior data includes the historical commodities and historical coupons corresponding to each behavior of the user.

[0148] The candidate coupon determination module is used to determine, for each of the historical commodities, the candidate coupons corresponding to the historical commodities based on the historical coupons, wherein the candidate coupons include the historical coupons whose similarity with the historical commodities is greater than or equal to the similarity threshold.

[0149] The storage module is used to associate and store the above historical commodities with the corresponding above candidate coupons.

[0150] The third construction module is used to determine the target historical product according to the historical product matching the pushed product, and to construct the coupon sample according to the candidate coupon associated with the target historical product.

[0151] In some embodiments, the coupon push device further includes:

[0152] The target candidate coupon determination module is used to determine the target candidate coupon based on the selection probability corresponding to the candidate coupon when the number of the candidate coupons associated with the target historical product is greater than 1, wherein the selection probability reflects the probability of the candidate coupon being selected by the user.

[0153] The fourth building module is used to build the coupon sample according to the target candidate coupon.

[0154] In some embodiments, the coupon push device further includes:

[0155] The matching degree calculation module is used to determine a first score and a second score for each of the candidate coupons associated with the target historical product based on the pushed product and the candidate coupon, wherein the first score reflects the price matching degree between the pushed product and the candidate coupon, and the second score reflects the matching degree between the consumption level of the user corresponding to the pushed product and the candidate coupon.

[0156] The selection probability calculation module is used to determine the selection probability corresponding to the candidate coupon according to the first score, the second score, the third score, the fourth score and the weights corresponding to the scores, the third score reflects the popularity of the candidate coupon, and the fourth score reflects the similarity between the candidate coupon and the target historical product.

[0157] In some embodiments, the coupon push device further includes:

[0158] The prediction module is used to use the user order data as the input of the pre-trained time prediction model to obtain the above-mentioned predicted redemption time output by the above-mentioned time prediction model. The above-mentioned user order data is used to reflect the user's commodity purchase status and coupon redemption status. The above-mentioned time prediction model is used to predict the time when the above-mentioned user redeems the above-mentioned coupon to be pushed based on the above-mentioned user order data to obtain the above-mentioned predicted redemption time.

[0159] In some embodiments, the time prediction model is also used to predict the time when the user collects the coupon to be pushed, and obtain the predicted collection time. The push module 24 includes:

[0160] The push unit is used to push the coupon to be pushed to the target user based on the corresponding predicted collection time.

[0161] In some embodiments, the coupon push device further includes:

[0162] The tag screening module is used to determine from various users the users whose user tags match the delivery tags corresponding to the coupons to be pushed, and obtain the screened users, wherein the user tags include at least one of a consumption level tag and an activity level tag.

[0163] Correspondingly, the above-mentioned basic user determination module 21 includes:

[0164] The screening unit is used to determine the above-mentioned basic users from the above-mentioned screened users according to the predicted verification time and the above-mentioned validity period.

[0165] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0166] Embodiment three:

[0167] Figure 3 This is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. Figure 3 As shown, the electronic device 3 of this embodiment includes: at least one processor 30 ( Figure 3 Only one processor is shown in the figure), a memory 31, and a computer program 32 stored in the memory 31 and executable on the at least one processor 30, wherein the processor 30 implements the steps in any of the above-mentioned method embodiments when executing the computer program 32.

[0168] The electronic device 3 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The electronic device may include, but is not limited to, a processor 30 and a memory 31. Those skilled in the art will appreciate that Figure 3 It is only an example of the electronic device 3 and does not constitute a limitation on the electronic device 3. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, it may also include input and output devices, network access devices, etc.

[0169] The processor 30 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0170] In some embodiments, the memory 31 may be an internal storage unit of the electronic device 3, such as a hard disk or memory of the electronic device 3. In other embodiments, the memory 31 may also be an external storage device of the electronic device 3, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device 3. Further, the memory 31 may also include both an internal storage unit of the electronic device 3 and an external storage device. The memory 31 is used to store an operating system, an application program, a boot loader (BootLoader), data, and other programs, such as the program code of the computer program, etc. The memory 31 may also be used to temporarily store data that has been output or is to be output.

[0171] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0172] An embodiment of the present application also provides a network device, which includes: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor implements the steps in any of the above-mentioned method embodiments when executing the computer program.

[0173] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0174] An embodiment of the present application provides a computer program product. When the computer program product runs on an electronic device, the electronic device can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0175] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device that can carry the computer program code to the camera / electronic device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, RandomAccess Memory), electric carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disk. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electric carrier signals and telecommunication signals.

[0176] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0177] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0178] In the embodiments provided in the present application, it should be understood that the disclosed devices / network equipment and methods can be implemented in other ways. For example, the device / network equipment embodiments described above are merely schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0179] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0180] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A coupon push method, characterized in that: include: Determine the basic user according to the predicted redemption time and the validity period of the coupon to be pushed, wherein the predicted redemption time is used to indicate the time when the user redeems the coupon to be pushed, and the basic user includes the user whose predicted redemption time matches the validity period; Determine, according to the user information of the basic user and the coupon information of the coupon to be pushed, the degree of matching between the basic user and the coupon to be pushed, wherein the degree of matching is used to reflect the interest degree of the basic user in the coupon to be pushed; Determine a target user from the basic users based on the matching degree; Push the coupon to be pushed to the target user.

2. The coupon push method according to claim 1, characterized in that: The determining, according to the user information of the basic user and the coupon information of the coupon to be pushed, a matching degree between the basic user and the coupon to be pushed includes: The user information of the basic user and the coupon information are used as inputs of a pre-trained analysis model to obtain the matching degree output by the analysis model. The analysis model is used to analyze the basic user's interest in the coupon to be pushed based on the user information and the coupon information to obtain the matching degree.

3. The coupon push method according to claim 2, characterized in that: Before using the user information of the basic user and the coupon information as inputs of the pre-trained analysis model, the method further includes: Acquire first user behavior data, where the first user behavior data includes user behavior data for pushing products and / or pushing coupons; In a case where the first user behavior data includes the pushed coupon, constructing a coupon sample based on the pushed coupon, And, in the case where the first user behavior data includes the pushed product, constructing the coupon sample according to the candidate coupon corresponding to the pushed product, the similarity between the candidate coupon and the pushed product is greater than or equal to a similarity threshold; The constructed analysis model is trained based on the constructed coupon sample and the corresponding user information to obtain the pre-trained analysis model.

4. The coupon push method according to claim 3, characterized in that: Before constructing the coupon sample according to the candidate coupons corresponding to the pushed products, the method further includes: Acquire second user behavior data corresponding to each user, wherein the second user behavior data includes historical commodities and historical coupons corresponding to each behavior of the user; For each of the historical commodities, determining the candidate coupons corresponding to the historical commodities based on the historical coupons, wherein the candidate coupons include the historical coupons whose similarity with the historical commodities is greater than or equal to the similarity threshold; storing the historical products and the corresponding candidate coupons in association with each other; Correspondingly, the step of constructing the coupon sample according to the candidate coupon corresponding to the pushed product includes: A target historical commodity is determined according to the historical commodity matching the pushed commodity, and the coupon sample is constructed according to the candidate coupon associated with the target historical commodity.

5. The coupon push method according to claim 4, characterized in that: The step of constructing the coupon sample according to the candidate coupon associated with the target historical commodity includes: When the number of the candidate coupons associated with the target historical product is greater than 1, determining a target candidate coupon based on a selection probability corresponding to the candidate coupons, wherein the selection probability reflects a probability that the candidate coupon is selected by the user; The coupon sample is constructed according to the target candidate coupon.

6. The coupon push method according to claim 5, characterized in that: Before determining the target candidate coupon based on the selection probability corresponding to the candidate coupon, the method further includes: For each candidate coupon associated with the target historical product, a first score and a second score are determined based on the pushed product and the candidate coupon, wherein the first score reflects the degree of price matching between the pushed product and the candidate coupon, and the second score reflects the degree of matching between the consumption level of the user corresponding to the pushed product and the candidate coupon; The selection probability corresponding to the candidate coupon is determined according to the first score, the second score, the third score, the fourth score and the weights corresponding to the scores, the third score reflects the popularity of the candidate coupon, and the fourth score reflects the similarity between the candidate coupon and the target historical product.

7. The coupon push method according to claim 1, characterized in that: Before determining the basic users according to the predicted redemption time and the validity period of the coupon to be pushed, the method further includes: The user order data is used as the input of the pre-trained time prediction model to obtain the predicted redemption time output by the time prediction model. The user order data is used to reflect the user's commodity purchase status and coupon redemption status. The time prediction model is used to predict the time when the user redeems the coupon to be pushed based on the user order data to obtain the predicted redemption time.

8. The coupon push method according to claim 7, characterized in that: The time prediction model is also used to predict the time when the user collects the coupon to be pushed, and obtain the predicted collection time. The pushing of the coupon to be pushed to the target user includes: The coupon to be pushed is pushed to the target user based on the corresponding predicted collection time.

9. The coupon push method according to any one of claims 1 to 8, characterized in that: Before determining the basic users according to the predicted redemption time and the validity period of the coupon to be pushed, the method further includes: Determine from each user a user whose user tag matches the delivery tag corresponding to the coupon to be pushed, and obtain the filtered user, wherein the user tag includes at least one of a consumption level tag and an activity level tag; Correspondingly, determining the basic users according to the predicted redemption time and the validity period of the coupon to be pushed includes: The basic users are determined from the screened users according to the predicted redemption time and the validity period.

10. A coupon push device, characterized in that: include: A basic user determination module, used to determine basic users according to a predicted redemption time and a validity period of the coupon to be pushed, wherein the predicted redemption time is used to indicate a time when a user redeems the coupon to be pushed, and the basic users include users whose predicted redemption time matches the validity period; A matching degree determination module, used to determine the matching degree between the basic user and the coupon to be pushed according to the user information of the basic user and the coupon information of the coupon to be pushed, wherein the matching degree is used to reflect the interest degree of the basic user in the coupon to be pushed; A target user determination module, configured to determine a target user from the basic users based on the matching degree; A push module is used to push the coupon to be pushed to the target user.

11. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the method according to any one of claims 1 to 9 is implemented.

12. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 9 is implemented.

Citation Information

Patent Citations

  • Categorized electronic coupon use prediction method and device and electronic equipment

    CN107330741A

  • Virtual article transmitting method and device, electronic equipment, and storage medium

    CN108416614A

  • Discount coupon issuing method and device, computer system and medium

    CN111242661A

  • Analysis method and device based on historical data and medium

    CN116151875A

  • Resource distribution method and device, storage medium and computer equipment

    CN116227826A