Electronic coupon issuing method, device and equipment, medium and program product

Abstract

The invention provides an electronic coupon issuing method which can be applied to the technical field of artificial intelligence. The electronic coupon issuing method comprises the following steps: for a pre-issued electronic coupon, under the condition of determining that the use condition of the issued electronic coupon meets a first preset condition, circularly executing the following steps until a preset stop condition is met: for a first user who has acquired the electronic coupon but does not use the electronic coupon currently, predicting the use probability of the electronic coupon; determining the number of the first users whose predicted use probabilities are lower than a preset probability lower limit as a reference number; determining the number of new users according to the reference number; for new users who do not acquire the electronic coupons, predicting a target user in the new users; the electronic coupon use probability of the target user is higher than that of the non-target user; issuing new electronic coupons for the target users whose number is the new user number; and triggering to execute the next cycle under the condition of determining that the new electronic coupon is issued and the use condition of the currently issued electronic coupon meets a second preset condition.

Description

Technical Field

[0001] This disclosure relates to the field of artificial intelligence technology, and more specifically to an electronic coupon issuance method, apparatus, device, medium, and program product. Background Technology

[0002] With the widespread use of the internet, more and more merchants are using e-coupons to conduct online marketing campaigns. For example, by issuing e-coupons or e-vouchers to users, they can offer discounts and thus encourage consumer spending.

[0003] However, when issuing e-coupons, they are often distributed to users randomly, which usually results in a large number of e-coupons being given to users who do not know how to use them, leading to waste of e-coupons and increased communication and storage costs.

[0004] Therefore, there is an urgent need for a method to improve the accuracy of electronic coupon distribution in order to reduce communication and storage costs. Summary of the Invention

[0005] In view of the above problems, this disclosure provides an electronic coupon distribution method, apparatus, equipment, medium and program product, which can improve the accuracy of electronic coupon distribution and reduce communication and storage costs.

[0006] According to a first aspect of this disclosure, a method for issuing electronic coupons is provided, comprising:

[0007] For pre-issued e-vouchers, if the usage of the issued e-vouchers meets the first preset condition, the following steps are executed repeatedly until the preset stop condition is met:

[0008] Based on a pre-trained e-voucher usage probability prediction model, the e-voucher usage probability is predicted for the first user who has obtained an e-voucher but has not yet used it; the number of first users whose predicted usage probability is lower than a preset probability lower limit is determined as the baseline number; and based on the determined baseline number, the number of new users for whom new e-vouchers need to be issued is determined.

[0009] Based on a pre-trained target user screening model, target users are predicted among new users who have not yet obtained e-coupons; the probability of e-coupon usage by target users is higher than that of non-target users.

[0010] New electronic coupons will be issued to target users whose number equals the number of new users.

[0011] Once it is confirmed that the issuance of new e-vouchers has been completed and the usage of currently issued e-vouchers meets the second preset condition, the next cycle will be triggered.

[0012] A second aspect of this disclosure provides an electronic coupon dispensing device, comprising:

[0013] The trigger module is used to trigger the loop module to perform operations when the usage of the pre-issued electronic coupons meets the first preset condition.

[0014] The loop module is used to repeatedly execute the following steps until a preset stopping condition is met:

[0015] Based on a pre-trained e-voucher usage probability prediction model, the e-voucher usage probability is predicted for the first user who has obtained an e-voucher but has not yet used it; the number of first users whose predicted usage probability is lower than a preset probability lower limit is determined as the baseline number; and based on the determined baseline number, the number of new users for whom new e-vouchers need to be issued is determined.

[0016] Based on a pre-trained target user screening model, target users are predicted among new users who have not yet obtained e-coupons; the probability of e-coupon usage by target users is higher than that of non-target users.

[0017] New electronic coupons will be issued to target users whose number equals the number of new users.

[0018] Once it is confirmed that the issuance of new e-vouchers has been completed and the usage of currently issued e-vouchers meets the second preset condition, the next cycle will be triggered.

[0019] A third aspect of this disclosure provides an electronic device comprising: one or more processors; and a memory for storing one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the method described above.

[0020] A fourth aspect of this disclosure also provides a computer-readable storage medium having a computer program or instructions stored thereon, which, when executed by a processor, implement the steps of the above-described method.

[0021] The fifth aspect of this disclosure also provides a computer program product, including a computer program or instructions that, when executed by a processor, implement the steps of the above-described method.

[0022] The above one or more embodiments have the following beneficial effects: by using a pre-trained electronic coupon usage probability prediction model and a target user screening model, electronic coupons are distributed to target users with a higher probability of using electronic coupons. Compared with the method of randomly distributing electronic coupons, this can improve the accuracy of electronic coupon distribution and reduce communication and storage costs.

[0023] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0024] The foregoing contents, as well as other objects, features, and advantages of this disclosure, will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0025] Figure 1 This illustration schematically depicts an application scenario of an electronic coupon issuance method and apparatus according to an embodiment of the present disclosure;

[0026] Figure 2 A flowchart illustrating an electronic coupon issuance method according to an embodiment of the present disclosure is shown schematically;

[0027] Figure 3 The diagram schematically illustrates an architecture of an intelligent marketing method according to an embodiment of the present disclosure;

[0028] Figure 4 A flowchart illustrating an intelligent marketing method according to an embodiment of the present disclosure is shown schematically;

[0029] Figure 5 This schematic diagram illustrates a structural block diagram of an electronic coupon issuing device according to an embodiment of the present disclosure;

[0030] Figure 6 A block diagram schematically illustrates an electronic device suitable for implementing an electronic coupon distribution method according to an embodiment of the present disclosure. Detailed Implementation

[0031] The embodiments of this disclosure will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of this disclosure. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of this disclosure for ease of explanation. However, it will be apparent that one or more embodiments may be practiced without these specific details. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concepts of this disclosure. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the disclosure. The terms “comprising,” “including,” etc., as used herein, indicate the presence of said features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components. All terms used herein (including technical and scientific terms) have the meaning commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted in a manner consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner. When using expressions such as "at least one of A, B, and C," the expression should generally be interpreted according to the meaning commonly understood by those skilled in the art (e.g., "a system having at least one of A, B, and C" should include, but is not limited to, systems having only A, only B, only C, A and B, A and C, B and C, and / or systems having A, B, and C, etc.). In the technical solutions of this disclosure, the user information (including but not limited to user personal information, user image information, user device information, such as location information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, storage, use, processing, transmission, provision, disclosure, and application of related data all comply with relevant laws, regulations, and standards, necessary confidentiality measures have been taken, and public order and good morals are not violated. Corresponding operation entry points are provided for users to choose to authorize or refuse. In scenarios involving automated decision-making using personal information, the methods, devices, and systems provided in the embodiments of this disclosure all provide users with corresponding operation entry points for users to choose to agree to or refuse the automated decision results; if the user chooses to refuse, the expert decision-making process is initiated. The term "automated decision-making" here refers to the activity of automatically analyzing and evaluating an individual's behavioral habits, interests, or economic, health, and credit status through computer programs, and then making decisions accordingly. The term "expert decision-making" here refers to the activity of making decisions by individuals who specialize in a particular field, possess specialized experience, knowledge, and skills, and have reached a certain level of professional expertise.

[0032] With the widespread use of the internet, more and more merchants are using e-coupons for online marketing campaigns. For example, issuing e-coupons or e-vouchers can offer discounts to users, thereby encouraging spending. However, e-coupons are often distributed randomly, frequently giving more coupons to users who don't know how to use them, resulting in wasted coupons and increased communication and storage costs. Therefore, there is an urgent need for a method to improve the accuracy of e-coupon distribution in order to reduce communication and storage costs.

[0033] This disclosure provides an electronic coupon distribution method that improves the accuracy of electronic coupon distribution and reduces communication and storage costs. In this method, the electronic coupon distribution process can be divided into multiple stages. Each stage can determine the number of electronic coupons to be distributed and the target users based on the electronic coupon usage data from previous stages. The electronic coupons are then distributed to the corresponding target users according to the required quantity. Target users are those with a high probability of using the electronic coupons, thereby increasing the utilization rate of the electronic coupons by distributing them to these target users.

[0034] Therefore, this method can improve the accuracy of electronic coupon distribution and increase the utilization rate of the distributed electronic coupons by distributing them to users with a higher probability of using them based on the usage of electronic coupons in the previous stage. This, in turn, can reduce the waste of communication and storage costs.

[0035] Figure 1 The illustration shows an application scenario of an electronic coupon issuance method and apparatus according to an embodiment of the present disclosure.

[0036] like Figure 1 As shown, application scenario 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 serves as a medium for providing a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired or wireless communication links, or fiber optic cables, etc.

[0037] Users can interact with server 105 via network 104 using first terminal device 101, second terminal device 102, and third terminal device 103 to receive or send messages, etc. Various communication client applications can be installed on first terminal device 101, second terminal device 102, and third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social media platform software, etc. (for example only). The client applications can receive electronic coupons issued by the corresponding server. First terminal device 101, second terminal device 102, and third terminal device 103 can be various electronic devices with displays and supporting web browsing, including but not limited to smartphones, tablets, laptops, and desktop computers. Server 105 can be a server providing various services, such as a backend management server supporting websites browsed by users using first terminal device 101, second terminal device 102, and third terminal device 103 (for example only). The backend management server can analyze and process received user requests and other data, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal devices.

[0038] The electronic coupon distribution method provided in this disclosure can generally be executed by server 105. Correspondingly, the electronic coupon distribution device provided in this disclosure can generally be located in server 105. The electronic coupon distribution method provided in this disclosure can also be executed by a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105. Correspondingly, the electronic coupon distribution device provided in this disclosure can also be located in a server or server cluster that is different from server 105 and capable of communicating with the first terminal device 101, the second terminal device 102, the third terminal device 103, and / or server 105. In one example, server 105 can execute the electronic coupon distribution method provided in this disclosure to distribute electronic coupons to different user accounts, specifically, to user terminal devices logged into the corresponding accounts.

[0039] This disclosure does not limit the application areas of the electronic coupon issuance method. The electronic coupon issuance method and apparatus specified in the embodiments of this disclosure can be used in the fintech field, as well as in any other field outside of fintech (e.g., shopping, catering, e-commerce, logistics, etc.). It should be understood that... Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0040] The following will be based on Figure 1 The described scene, through Figures 2-6 A method for issuing electronic coupons according to an embodiment of this disclosure will be described in detail. Figure 2 A flowchart illustrating an electronic coupon distribution method according to an embodiment of the present disclosure is shown schematically. Figure 2 As shown, an electronic coupon issuance method according to an embodiment of this disclosure may include operations S210 to S250. This disclosure does not limit the executing entity of the electronic coupon issuance method; it can be executed by any electronic device, such as a server.

[0041] In operation S210, for pre-issued electronic coupons, if it is determined that the usage of the issued electronic coupons meets the first preset condition, operations S220 to S250 are executed cyclically until the preset stop condition is met.

[0042] In operation S220, based on the pre-trained electronic coupon usage probability prediction model, the electronic coupon usage probability is predicted for the first user who has obtained an electronic coupon but has not yet used it; the number of first users whose predicted usage probability is lower than the preset probability lower limit is determined as the baseline number; based on the determined baseline number, the number of new users who need to be issued new electronic coupons is determined.

[0043] In operation S230, based on a pre-trained target user screening model, target users are predicted among new users who have not obtained electronic coupons.

[0044] Among them, the probability of electronic coupons being used by target users is higher than that of non-target users.

[0045] In operation S240, new electronic coupons are issued to target users whose number is equal to the number of new users.

[0046] In operation S250, if it is determined that the issuance of new electronic coupons has been completed and the usage of currently issued electronic coupons meets the second preset condition, the next cycle is triggered.

[0047] Specifically, triggering the execution of the next cycle in operation S250 can be triggered to re-execute operations S220 to S250.

[0048] This method uses a pre-trained e-voucher usage probability prediction model and a target user screening model to predict the number of e-vouchers to be issued in the current cycle and the target users to be issued to. E-vouchers are then issued to target users with a higher probability of e-voucher usage. Compared with random e-voucher issuance, this method can improve the accuracy of e-voucher issuance, increase the proportion of users who use e-vouchers, reduce the proportion of users who do not use e-vouchers, and increase the user utilization rate of issued e-vouchers. This can reduce the waste of communication and storage costs.

[0049] This method can also expand the reach of e-coupons to users, improve the matching degree between users and e-coupons, reduce the proportion of users who do not use e-coupons, and increase the usage rate of issued e-coupons. This can reduce the disturbance to users caused by issuing e-coupons and improve the user experience.

[0050] This method does not limit the specific execution order of the steps. Optionally, S220 and S230 can be executed in parallel or sequentially. Optionally, S240 and S250 can be executed after S220 and S230.

[0051] In operations S220 and S230, user information can be used for model prediction. Therefore, in embodiments of this disclosure, user consent or authorization can be obtained before acquiring user information. For example, a request to acquire user information can be sent to the user before operation S220. Operation S220 is executed if the user consents or authorizes the acquisition of user information. The same approach can be used for other steps that require the use of user information.

[0052] In the embodiments of this disclosure, a corresponding operation entry point can be provided to the user, allowing the user to choose to agree to or reject the automated decision result. For example, before processing / making a decision on user information in operation S220, the user can provide an instruction to agree to or reject the aforementioned processing / decision through the corresponding operation entry point. If the user agrees to the aforementioned processing / decision, then operation S220 is performed on the user information, i.e., operation S220 is executed. If the user refuses to perform the aforementioned processing / decision, then the expert decision-making process is initiated. Similarly, the same approach can be used for operation S230. For other steps in this method flow that also use user information, the same approach can be used to protect user information security.

[0053] I. Explanation of Operation S210.

[0054] In operation S210, for pre-issued electronic coupons, if it is determined that the usage of the issued electronic coupons meets the first preset condition, operations S220 to S250 are executed cyclically until the preset stop condition is met.

[0055] This methodology does not limit the specific form and content of the e-voucher. Optionally, the e-voucher can be a virtual resource, such as a coupon, voucher, or discount coupon, which can be used for promotional marketing to attract more users.

[0056] It should be noted that the electronic coupons mentioned in different embodiments of this method can be the same specific type of electronic coupon, such as coupons issued in a single marketing campaign; discount coupons issued for the same product; vouchers issued by the same organization, etc. Multiple electronic coupons can be issued to a single user, or only one electronic coupon can be issued to a single user. Electronic coupons can be exchanged or transferred between users, or their exchange or transfer between users can be prohibited. The same electronic coupon can be used for multiple redemptions, or it can be restricted to a single redemption.

[0057] This method distinguishes different stages of e-coupon distribution through a loop, distributing e-coupons separately at each stage. Before the first loop, a portion of e-coupons can be pre-distributed to determine their usage and facilitate subsequent operations.

[0058] This method does not limit the quantity or method of pre-issuing e-vouchers. Optionally, the quantity of e-vouchers to be pre-issued can be determined based on business needs. For example, based on the marketing budget in the business needs, the total number of e-vouchers that can be issued can be determined, and thus the quantity of e-vouchers that need to be pre-issued can be determined according to a preset ratio.

[0059] Optionally, when issuing electronic coupons in advance, electronic coupons can be issued to users randomly; alternatively, users with a higher probability of using electronic coupons can be selected based on a pre-trained user screening model. Specifically, users whose probability of using electronic coupons is higher than a preset probability limit can be selected, and electronic coupons can be issued to the selected users.

[0060] This method does not limit the specific content of the first preset condition. Optionally, it could be that the usage rate of the pre-issued e-vouchers exceeds a specified usage rate threshold, thereby triggering an operation to issue additional e-vouchers. Alternatively, it could be that from the moment the e-vouchers are pre-issued, after a specified period of time, the operation can be triggered to issue additional e-vouchers, specifically, e-vouchers can be issued at fixed intervals.

[0061] Therefore, the first preset condition may optionally include at least one of the following:

[0062] 1) The percentage of users who have obtained and used electronic coupons among all users who have obtained electronic coupons is greater than the first percentage threshold.

[0063] 2) The duration between the issuance time of the targeted electronic coupon and the current time is greater than the first duration threshold.

[0064] This embodiment can improve the efficiency of issuing electronic coupons and expand the user reach of electronic coupons by limiting specific first preset conditions and triggering the issuance of new electronic coupons when the usage rate or duration meets the preset conditions. This embodiment does not limit the specific method for determining the issuance time of electronic coupons. Optionally, since multiple electronic coupons can be issued simultaneously or not simultaneously, if the electronic coupons issued in advance are issued simultaneously, the time when the electronic coupons are issued simultaneously can be determined as the issuance time of the electronic coupons. If the electronic coupons issued in advance are not issued simultaneously, the earliest time, latest time, or intermediate time when the electronic coupons are issued can be determined as the issuance time of the electronic coupons. Furthermore, this embodiment does not limit the specific values ​​of the first proportion threshold and the first duration threshold.

[0065] Optionally, operations S220 to S250 can be triggered cyclically if the usage of the issued electronic coupon meets any one of multiple different first preset conditions, until a preset stop condition is met. It is understood that, for multiple different first preset conditions, operations S220 to S250 can be triggered cyclically if any one or more of the first preset conditions are met, until the preset stop condition is met.

[0066] This method does not limit the specific way of determining whether the first preset condition is met. Optionally, the usage of issued electronic coupons can be checked periodically or continuously to determine whether the first preset condition is met.

[0067] This method does not limit the specific way operations S220 to S250 are executed in a loop. It is understood that the next iteration of the loop can be triggered by an operation within the loop, specifically by operation S250. Optionally, operation S250 can continuously monitor the issuance of new e-coupons and the usage of currently issued e-coupons.

[0068] This method does not limit the preset stopping conditions for the loop. Optionally, the loop can stop after a specified duration, after a specified number of loops, after the current time exceeds a specified time, or depending on the actual scenario, the loop can stop after the total number of electronic coupons issued exceeds a specified total number.

[0069] Therefore, optionally, the preset stop condition may include at least one of the following: 1) the duration of the loop is greater than the preset duration threshold; 2) the number of loops is greater than the preset number of loops threshold; 3) the current time exceeds the preset stop time; 4) the total number of electronic coupons currently issued is greater than the preset total number threshold.

[0070] In a specific example, the issued e-vouchers can be vouchers that can be used to offset a certain amount. In actual business operations, it's necessary to control the total amount that can be offset by all e-vouchers to facilitate cost control. Therefore, a preset threshold can be set for the total number of e-vouchers that can be issued under a specified marketing cost, thus controlling the total number of e-vouchers issued and facilitating cost management. When marketing costs are similar, compared to randomly issuing e-vouchers, this example can reduce wasted marketing costs, increase e-voucher utilization, and enhance marketing effectiveness.

[0071] In another specific example, the e-voucher may be issued for a specific marketing campaign, and the validity period or expiration date of the e-voucher can be set, which may be the duration or expiration date of the marketing campaign, thereby setting a corresponding preset duration threshold or preset stop time.

[0072] This method does not limit the specific way of determining whether the preset stopping condition is met. Optionally, it can periodically check whether the preset stopping condition is met, or it can continuously monitor whether the preset stopping condition is met. It is understood that the determination of whether the preset stopping condition is met can be made during the loop execution. If the preset stopping condition is determined to be met, the above loop operation can be stopped directly, without waiting for the current loop execution to complete.

[0073] II. Explanation of Operation S220.

[0074] In operation S220, based on a pre-trained e-coupon usage probability prediction model, the e-coupon usage probability is predicted for the first user who has obtained an e-coupon but has not yet used it. The number of first users whose predicted usage probability is lower than a preset probability lower limit is determined as the baseline number. Based on the determined baseline number, the number of new users for whom new e-coupons need to be issued is determined. For ease of description, users who have obtained e-coupons but have not yet used them are referred to as the first users. The new e-coupon can be the same type as the previously issued e-coupons; for ease of description, the e-coupons newly issued during the cycle are referred to as new e-coupons.

[0075] 1. Regarding the prediction model for the probability of using electronic coupons.

[0076] This method does not limit the specific structure of the e-coupon usage probability prediction model. Optionally, the e-coupon usage probability prediction model can be a random forest model or a neural network model. Optionally, the e-coupon usage probability prediction model can be used to predict the probability of e-coupon usage for users.

[0077] This methodology does not limit the training method or timing of the electronic coupon usage probability prediction model. Specific training methods can include supervised training, unsupervised training, or semi-supervised training, etc. Optionally, the electronic coupon usage probability prediction model can be pre-trained before executing this methodology; it can also be pre-trained before executing the iterative steps.

[0078] Regarding the inputs and outputs of the electronic coupon usage probability prediction model, this method does not limit the specific form.

[0079] Optionally, the input to the e-coupon usage probability prediction model can be used to characterize relevant user information, specifically information that can be used to predict the probability of e-coupon usage. This embodiment does not limit the specific content of the characterized user-related information. Optionally, the user-related information may include the user's basic information and / or e-coupon-related information. For example, the user's basic information may include the user's gender, age, spending habits, transaction history, etc. The e-coupon-related information may include the method of obtaining the e-coupon, the duration of obtaining the e-coupon, the time of obtaining the e-coupon, the number of times the user viewed the e-coupon, etc. This user-related information may be obtained with the user's authorization.

[0080] Optionally, the output of the e-coupon usage probability prediction model can be used to characterize the probability of a user using an e-coupon.

[0081] Furthermore, the sample features of the training samples for the electronic coupon usage probability prediction model can be matched with the input of the model, specifically, they can have the same form. This method does not limit the form of the training samples for the electronic coupon usage probability prediction model. Optionally, the training samples for the electronic coupon usage probability prediction model may include: sample features representing user-related information, and sample labels representing the user's electronic coupon usage probability.

[0082] This embodiment does not limit the specific content and form of the sample features. Optionally, the sample features in the training samples of the electronic coupon usage probability prediction model may include at least one of the following: user information and electronic coupon information. User information may include at least one of the following: basic user information, historical transaction information, and current electronic coupon usage status, etc. Among them, basic user information may include at least one of the following: user gender, user place of origin, user age, user ethnicity, user region, user occupation, etc. Electronic coupon information may include at least one of the following: electronic coupon acquisition method, electronic coupon acquisition time, number of times the user has viewed the electronic coupon, etc. Historical transaction information may include at least one of the following: number of historical transactions, amount of a single historical transaction, total amount of historical transactions, type of historical transactions, etc. Current electronic coupon usage status may include at least one of the following: whether the electronic coupon has been used, the time of electronic coupon usage, the amount of a single transaction using the electronic coupon, the type of transaction using the electronic coupon, the total number of transactions using the electronic coupon (multiple redemptions are allowed), the total amount of transactions using the electronic coupon, the time between acquisition and use of the electronic coupon, etc.

[0083] This embodiment does not limit the specific content and format of the sample labels. Optionally, the sample labels in the training samples of the e-coupon usage probability prediction model can be used to represent that the user's e-coupon usage probability is 0 or 1. Specifically, when labeling training samples, the corresponding e-coupon usage probability can be determined based on whether the user uses the e-coupon. For users who use e-coupons, the corresponding training sample can be labeled as "e-coupon usage probability is 1"; for users who do not use e-coupons, the corresponding training sample can be labeled as "e-coupon usage probability is 0". Of course, other probability values ​​can also be used for labeling, such as "e-coupon usage probability is 40%".

[0084] This embodiment does not limit the source of training samples for the e-coupon usage probability prediction model. Optionally, training samples can be constructed by collecting relevant user data based on previous user usage of e-coupons. Alternatively, training samples can be constructed by collecting relevant user data from users who have obtained e-coupons as the loop steps are executed.

[0085] For example, in the third iteration, training samples can be built using user data from users who used e-vouchers in the first two iterations, for further training of the e-voucher usage probability prediction model.

[0086] This method does not limit the way training samples are constructed for the e-coupon usage probability prediction model. Optionally, the training samples can be constructed by filling in relevant information and data based on the form of sample features and sample labels, according to the user's relevant information and the user's e-coupon usage or e-coupon usage probability.

[0087] 2. Regarding training optimization.

[0088] In an alternative embodiment, in conjunction with the looping steps, in any loop (e.g., in the first loop), the probabilistic prediction model can be further optimized and trained for the current electronic voucher before continuing to execute operation S220.

[0089] Alternatively, in each iteration (or multiple iterations), the current electronic coupon usage probability prediction model can be further optimized and trained to improve the training effect and prediction accuracy of the electronic coupon usage probability prediction model.

[0090] Therefore, optionally, based on the pre-trained electronic coupon usage probability prediction model, the electronic coupon usage probability is predicted for the first user who has obtained an electronic coupon but has not yet used it. Specifically, this can be done by: optimizing and training the current electronic coupon usage probability prediction model; and based on the optimized and trained current electronic coupon usage probability prediction model, predicting the electronic coupon usage probability for the first user who has obtained an electronic coupon but has not yet used it.

[0091] This embodiment can improve the training effect and prediction accuracy of the electronic coupon usage probability prediction model by further optimizing and training the model in each iteration.

[0092] In the first iteration, the current e-coupon usage probability prediction model before optimization training can be either the initial e-coupon usage probability prediction model or a trained e-coupon usage probability prediction model. In subsequent iterations, the current e-coupon usage probability prediction model before optimization training can be the e-coupon usage probability prediction model obtained through optimization training in previous iterations.

[0093] This embodiment does not limit the specific method of optimization training. Optionally, new training samples can be added and optimization training can be performed based on the new training samples; alternatively, the model structure or loss function can be adjusted before optimization training.

[0094] In one alternative embodiment, since new e-coupons may be issued during the loop, new training samples can be constructed or updated to obtain new training samples based on the user data of users who have already obtained e-coupons in each loop, and the e-coupon usage probability prediction model can be optimized and trained based on the newly added training samples.

[0095] By executing the loop multiple times, the number of training samples can be increased, and the accuracy of the sample labels can be improved. With the optimization training in each loop, the training effect and prediction accuracy of the electronic coupon usage probability prediction model can be improved.

[0096] Therefore, optionally, the current electronic coupon usage probability prediction model can be optimized and trained by: constructing a first training sample set based on the user data of users who have obtained electronic coupons; optimizing and training the current electronic coupon usage probability prediction model based on the first training sample set; in any training sample of the first training sample set, the sample features are used to represent user information and electronic coupon information, and the sample label is used to represent that the user's electronic coupon usage probability is 0 or 1.

[0097] This embodiment can further optimize the training of the current electronic coupon usage probability prediction model by constructing training samples based on the user data of users who have obtained electronic coupons in each loop. In this way, the training effect and prediction accuracy of the electronic coupon usage probability prediction model can be improved by combining the loop steps and optimizing the training by increasing the number of training samples with each loop.

[0098] This embodiment does not limit the specific method of constructing the first training sample set. Optionally, training samples can be directly constructed and added to the first training sample set; alternatively, the first training sample set can be updated based on the updated user data of users who have already obtained e-coupons, thereby improving the efficiency of constructing the first training sample set. For example, for users who have previously obtained new e-coupons, their user data can be added, thus creating new training samples that are added to the first training sample set. For users who have not previously used e-coupons, if they use e-coupons after a period of time, their user data can be updated, and corresponding training samples can be added or updated to the first training sample set.

[0099] This embodiment does not limit the specific optimization training method. Optionally, the entire first training sample set can be used for optimization training in each cycle of optimization training; alternatively, a portion of the training samples can be sampled from the first training sample set for optimization training in each cycle of optimization training. Specifically, this can involve sampling newly added or updated training samples, thereby improving the efficiency of optimization training.

[0100] This method does not limit the specific way of constructing training samples. The training samples here can be used to train the electronic coupon usage probability prediction model.

[0101] Optionally, the construction method of the training samples for the e-coupon usage probability prediction model may specifically include: among users who have currently obtained e-coupons, for any user who has obtained and used an e-coupon, constructing a training sample based on the user data of the target user; the sample label of the constructed training sample is used to represent that the user's e-coupon usage probability is 1; among users who have currently obtained e-coupons, for any user who has obtained but not used an e-coupon, constructing a training sample with a pseudo-label based on the user data of the target user; the pseudo-label of the constructed training sample is used to represent that the user's e-coupon usage probability is 0.

[0102] This embodiment can construct corresponding positive samples and pseudo-labeled negative samples based on the user's electronic coupon usage, thereby reducing the difficulty of obtaining and constructing training samples, increasing the number of training samples, and improving the model training effect.

[0103] The users who have already obtained e-coupons can include those who received pre-issued e-coupons, as well as new users who obtain new e-coupons as the loop continues. This allows for the construction of training samples by combining user data and e-coupon usage patterns. By using recent user data to construct training samples and defining sample labels based on e-coupon usage, the constructed training samples are more closely aligned with the current e-coupon scenario, improving the adaptability of the e-coupon usage probability prediction model to the current e-coupon situation.

[0104] The sample features of the constructed training samples can be used to represent relevant user information. For positive samples, since the samples are constructed for users who have used e-coupons, the corresponding sample labels can be determined as sample labels representing a user's e-coupon usage probability of 1. For negative samples, since we are currently in a loop, for users who have not used e-coupons, we can temporarily assume that these users will not use e-coupons in this loop, or that these users will not use e-coupons, thus obtaining corresponding pseudo-labels representing a user's e-coupon usage probability of 0. Understandably, as the loop continues to execute, some pseudo-labeled negative samples can be updated to positive samples.

[0105] This embodiment explains the construction of positive samples and pseudo-labeled negative samples. It can be understood that negative samples can also be constructed directly, specifically by constructing negative samples from historical user data of users who did not use the e-coupon until it expired.

[0106] This embodiment is for illustrative purposes only. It is understood that this method does not limit the format or meaning of the training sample labels. Optionally, the sample labels of the constructed training samples can be used to represent specific electronic coupon usage probability values, such as 40%, 60%, etc.

[0107] This embodiment explains how to construct a single sample. It is understood that, in order to construct multiple samples, corresponding samples can be constructed for different users using this embodiment.

[0108] This method does not limit the specific users used to construct the samples. Optionally, a subset of users can be sampled or selected from those who have already obtained electronic coupons to construct the training samples. This embodiment does not limit the specific sampling or selection method.

[0109] Optionally, since users who have already acquired and used e-coupons can be used to construct positive samples, positive samples can be constructed directly based on all or some of the users who have already acquired and used e-coupons.

[0110] Optionally, users who have currently acquired but not used e-coupons can be used to construct pseudo-label negative samples. Pseudo-label negative samples can be constructed directly based on all or some of the users who have currently acquired but not used e-coupons. To improve the accuracy of the negative samples, users whose e-coupons have expired and not been used can be further identified and their negative samples constructed.

[0111] In this method, the electronic coupons issued in advance and the new electronic coupons issued during the loop can be the same type of electronic coupon, thus expiring at the same time later. Alternatively, different electronic coupons can have separate validity periods, expiring after the validity period begins from the time of issuance. Therefore, for users who have not used their electronic coupons, some users' electronic coupons may have already expired, thus determining that the probability of using the electronic coupon is 0.

[0112] To improve the accuracy of negative samples, cluster analysis can be performed on users who have already obtained e-coupons. Specifically, the number of clusters can be set to two or other values. The clustering results can be analyzed: for any cluster, if it contains a large number of users who have used e-coupons, the probability of e-coupon use in that cluster is considered high; conversely, if it contains a small number of users who have used e-coupons, the probability of e-coupon use in that cluster is considered low. Furthermore, for the clusters with low e-coupon use probabilities, users who have not used e-coupons can be selected to construct pseudo-label negative samples.

[0113] The constructed labeled samples can be added to the first training sample set to optimize the training probability model for e-voucher usage. The method for constructing the training samples in the first training sample set can refer to the training sample construction method described above.

[0114] 3. Regarding the baseline number and the number of new users.

[0115] After training the e-voucher usage probability prediction model, you can use the trained e-voucher usage probability prediction model to predict the probability of e-voucher usage for each user.

[0116] In one alternative embodiment, for users who have already used e-coupons, there is no need to predict the e-coupon usage probability. However, for users who have not yet used e-coupons, since subsequent cyclical steps need to be performed, the e-coupon usage probability can be predicted to determine whether they will use the e-coupons in the future. This can help determine whether e-coupons need to be issued to new users and the number of new users to be issued to.

[0117] Optionally, based on a pre-trained e-coupon usage probability prediction model, the probability of e-coupon usage can be predicted for the first user who has obtained an e-coupon but has not yet used it. It is understood that users with a low predicted e-coupon usage probability can be considered as users who will not use e-coupons in the future. Therefore, the number of first users whose predicted usage probability is below a preset probability lower limit can be determined as a baseline number. This baseline number can be the number of users among the predicted users who will not use e-coupons in the future, and can be used to determine the number of users for whom new e-coupons need to be issued.

[0118] This embodiment does not limit the number of first users predicted. Optionally, it can be based on an e-coupon usage probability prediction model to predict the probability of e-coupon usage for all or part of the first users, and determine the corresponding baseline number. Therefore, the baseline number can be the number of users among the predicted users who do not subsequently use the e-coupon. Optionally, it can be based on a pre-trained e-coupon usage probability prediction model to predict the e-coupon usage probability for all first users who have obtained e-coupons but have not yet used them.

[0119] This embodiment does not limit the specific value of the preset probability lower limit. Optionally, the preset probability lower limit can be 10% or 15%. The preset probability lower limit can also be determined in other ways, such as by determining it based on the predicted probability distribution of electronic coupon usage, etc.

[0120] This method does not limit the specific way of determining the number of new users based on the baseline quantity. Optionally, the baseline quantity can be directly determined as the number of new users. That is, for users who will not use the e-coupons in the future, the same number of new users can be issued e-coupons to expand the user reach of the e-coupons and increase the usage rate of the e-coupons.

[0121] Optionally, the number of new users can be determined by considering the current usage of e-coupons. Based on the established baseline number, the number of new users to be issued e-coupons can be determined. Specifically, this can be done by: determining the number of users who have already obtained and used e-coupons, and the percentage of users who have already obtained e-coupons; and then determining the number of new users to be issued e-coupons based on the ratio between the established baseline number and the percentage of users who have already used e-coupons.

[0122] This embodiment combines the current user usage rate of e-coupons with the predicted number of users who will not use e-coupons in the future to determine the number of new users for whom new e-coupons need to be issued. This can increase the user reach of e-coupons and improve their usage rate. Specifically, the user usage ratio can be calculated by dividing the number of users who have currently obtained and used e-coupons by the total number of users who have obtained e-coupons.

[0123] This embodiment does not limit the specific method for determining the number of new users. Optionally, the determined baseline number can be divided by the user usage percentage, and the result rounded down to obtain the number of new users. Alternatively, the baseline number can be divided by the user usage percentage, the result rounded down further, and then a preset error can be added or subtracted to obtain the number of new users. This preset error can be set based on experience or data analysis.

[0124] This method does not limit the number or form of the issued e-vouchers. Optionally, multiple e-vouchers can be issued to any one user at a time; alternatively, only one e-voucher can be issued to different users, with a restriction that each user can only hold one e-voucher. In a specific example, if only one e-voucher can be issued to any one user, the number of new e-vouchers to be issued can be determined based on a baseline quantity. For a more detailed explanation, please refer to other embodiments.

[0125] III. Regarding operation S230.

[0126] In operation S230, based on a pre-trained target user screening model, target users are predicted among new users who have not yet obtained electronic coupons.

[0127] The probability of target users using e-coupons can be higher than that of non-target users. Understandably, the target user screening model can be used to identify target users with a higher probability of using e-coupons, which can then be used to determine the users to whom new e-coupons are subsequently issued. Therefore, target users can specifically be users with a high probability of using e-coupons, users whose probability of using e-coupons exceeds a preset probability upper limit, or users whose probability of using e-coupons exceeds that of non-target users (other users). Specifically, the target user screening model can be a binary classification model, which can predict whether a user is a target user or a non-target user.

[0128] 1. Regarding the target user screening model.

[0129] This method does not limit the specific structure of the target user screening model. Optionally, the target user screening model can employ logistic regression, random forest, neural network models, etc. Optionally, the target user screening model can be used to predict whether a user belongs to the target user or not, specifically it can be a binary classification model.

[0130] In a specific example, the target user screening model can also be used to predict the probability of e-coupon usage. Based on the predicted usage probability, a threshold can be set to predict users with a usage probability higher than the threshold as target users and users with a usage probability lower than the threshold as non-target users. Alternatively, a threshold range including an upper and lower limit can be set, predicting users with a usage probability higher than the upper limit as target users and users with a usage probability lower than the lower limit as non-target users.

[0131] For an explanation of the target user screening model, please refer to other embodiments.

[0132] The target user selection model differs from the e-coupon usage probability prediction model described above. The e-coupon usage probability prediction model can be used to predict the users who have already obtained e-coupons, specifically for the first user. The target user selection model, however, needs to predict the users who have not yet obtained e-coupons, thus facilitating the identification of users who need to receive e-coupons. Correspondingly, the input formats of the models can also differ. The input to the e-coupon usage probability prediction model can include e-coupon information, while the input to the target user selection model should not include e-coupon information, because the user data for new users who have not yet obtained e-coupons typically does not include e-coupon information.

[0133] This method does not limit the specific training method or timing of the target user selection model. The specific training method can be supervised training, unsupervised training, or semi-supervised training, etc. Optionally, the target user selection model can be pre-trained before executing this method, so that the pre-trained target user selection model can be directly used in operation S230; alternatively, the target user selection model can be pre-trained before executing the loop steps.

[0134] This method does not limit the specific form of the input and output of the target user selection model.

[0135] Optionally, the input to the target user screening model can be used to characterize relevant user information, specifically information that can be used to predict whether a user belongs to the target user group. This embodiment does not limit the specific content of the characterized user-related information. Optionally, the user-related information may include the user's own basic information.

[0136] Optionally, the output of the target user screening model can be used to characterize whether a user belongs to the target user, or it can be used to characterize the probability that a user belongs to the target user.

[0137] This method does not limit the form of training samples for the target user to select the model.

[0138] Optionally, the training samples of the target user screening model may include: sample features used to characterize user-related information, and sample labels used to characterize whether a user belongs to the target user.

[0139] This embodiment does not limit the specific content and form of the sample features. Optionally, the sample features in the training samples of the target user screening model may include at least one of the following: user information and transaction information. See other embodiments for details.

[0140] This embodiment does not limit the specific content and format of the sample labels. Optionally, the sample labels in the training samples of the target user screening model can be used to characterize whether a user belongs to the target user or not. The specific label format can be represented by 1 and 0 respectively.

[0141] This embodiment does not limit the source of training samples for the target user screening model. Optionally, relevant user data can be collected to construct training samples. See other embodiments for details.

[0142] This method does not limit the way training samples are constructed for the target user screening model. Optionally, the training samples can be constructed by filling in relevant information and data based on user information and the user's e-coupon usage history or probability, using sample features and sample labels. Specifically, the user's e-coupon usage history or probability can be used to determine whether the user is a target user, thereby determining the sample labels for the training samples.

[0143] 2. Regarding training optimization.

[0144] For an explanation of the optimized training of the target user screening model, please refer to other embodiments, specifically the optimized training embodiment of the electronic coupon usage probability prediction model described above.

[0145] In one optional embodiment, in conjunction with the looping steps, further optimization training can be performed on the current target user selection model in any loop (e.g., in the first loop) before executing S230. Alternatively, further optimization training can be performed on the current target user selection model in each loop (or multiple loops), thereby further improving the training effect and prediction accuracy of the target user selection model.

[0146] Therefore, optionally, based on the pre-trained target user screening model, for new users who have not obtained electronic coupons, the target users can be predicted. Specifically, this can be done by: optimizing the current target user screening model; and based on the optimized current target user screening model, predicting the target users for new users who have not obtained electronic coupons.

[0147] This embodiment can improve the training effect and prediction accuracy of the target user selection model by further optimizing the training of the current target user selection model in each loop.

[0148] In the first iteration, the current target user selection model before optimization training can be either the initial target user selection model or the target user selection model that has already been trained. In subsequent iterations, the current target user selection model before optimization training can also be the target user selection model obtained through optimization training in previous iterations.

[0149] This embodiment does not limit the specific method of optimization training. Optionally, new training samples can be added and optimization training can be performed based on the new training samples; the model structure can be adjusted and then optimization training can be performed; the loss function can be adjusted and then optimization training can be performed, etc.

[0150] Optionally, the current target user screening model can be optimized by: constructing a second training sample set based on the user data of users who have already obtained electronic coupons; optimizing the current target user screening model based on the second training sample set; in any training sample of the second training sample set, sample features are used to represent user information, and sample labels are used to represent whether the user is a target user.

[0151] This embodiment can further optimize the current target user selection model by constructing training samples based on user data of users who have already obtained electronic coupons in each iteration. This allows for optimization training by increasing the number of training samples with each iteration, thereby improving the training effect and prediction accuracy of the target user selection model. Furthermore, the accuracy of sample labels can be improved with each iteration to further enhance the training effect and prediction accuracy of the target user selection model. For a more detailed explanation, please refer to other embodiments.

[0152] This embodiment does not limit the specific method of constructing the second training sample set, nor does it limit the specific method of optimizing the training. Please refer to other embodiments for details.

[0153] This method does not limit the specific way of constructing training samples. The training samples here can be used to train the target user selection model. Specifically, the probability of e-coupon usage can be determined by considering users' e-coupon usage history. Users who have already used e-coupons can be identified as target users, i.e., users with a high probability of usage. Users who have not yet used e-coupons can be identified as non-target users, i.e., users with a low probability of usage; pseudo-labels can be set for this. It can be assumed that the probability of e-coupon usage for users who have already used e-coupons is higher than that for users who have not yet used e-coupons.

[0154] Optionally, the method for constructing the training samples of the target user screening model may specifically include: among users who have currently obtained electronic coupons, for any user who has obtained and used an electronic coupon, constructing a training sample based on the user data of the target user; the sample label of the constructed training sample is used to characterize that the user belongs to the target user; among users who have currently obtained electronic coupons, for any user who has obtained but not used an electronic coupon, constructing a training sample with pseudo-labels based on the user data of the target user; the pseudo-labels of the constructed training sample are used to characterize that the user belongs to a non-target user.

[0155] This embodiment can construct corresponding positive samples and pseudo-labeled negative samples based on the user's electronic coupon usage, thereby reducing the difficulty of obtaining and constructing training samples, increasing the number of training samples, and improving the model training effect.

[0156] The users who have already obtained e-coupons can include those who received pre-issued e-coupons, as well as new users who obtain new e-coupons as the loop continues. This allows for the construction of training samples by combining user data and e-coupon usage patterns. By using recent user data to construct training samples and defining sample labels based on e-coupon usage, the constructed training samples are more closely aligned with the current e-coupon scenario, improving the adaptability of the e-coupon usage probability prediction model to the current e-coupon situation.

[0157] The sample features of the constructed training samples can be used to represent relevant user information. For positive samples, since the samples are constructed for users who have used e-coupons, the corresponding sample labels can be determined as sample labels representing that the user belongs to the target user group. For negative samples, for users who have not currently used e-coupons, it can be temporarily assumed that these users will not use e-coupons in this loop, or that these users will not use e-coupons, thus obtaining corresponding pseudo-labels to represent that the user belongs to a non-target user group. It is understood that as the loop continues to execute, some pseudo-label negative samples can be updated to positive samples. This embodiment does not limit the construction method of negative samples or the construction method of multiple samples; please refer to other embodiments for details.

[0158] This embodiment explains how to construct a single sample. It is understood that, in order to construct multiple samples, corresponding samples can be constructed for different users using this embodiment.

[0159] This method does not limit the specific users used to construct the samples. Optionally, a subset of users can be sampled or selected from those who have already obtained e-coupons to construct the training samples. This embodiment does not limit the specific sampling or selection method. For specific sample construction methods, please refer to other embodiments. Optionally, users whose e-coupons have expired and not been used can be identified to construct negative samples. Cluster analysis can also be performed on users who have already obtained e-coupons to facilitate the construction of pseudo-labeled negative samples. It is understood that the constructed labeled samples can be added to the second training sample set to optimize the target user selection model. The construction method of the training samples in the second training sample set can refer to the above-described training sample construction method.

[0160] 3. Regarding the prediction of target users.

[0161] After obtaining a pre-trained (or optimized) target user selection model, the method can further predict target users among new users who have not yet received e-coupons based on this model. This method does not limit the number of predicted new users. Optionally, the method can predict target users among multiple new users who have not yet received e-coupons based on the pre-trained target user selection model. The number of predicted new users can be greater than the number of new users determined in operation S220. This method also does not limit the method of determining new users; specifically, new users can be randomly selected for prediction. Then, based on the prediction results of the target user selection model, target users can be determined among the predicted users. These target users can be considered new users with a high probability of using e-coupons, allowing new e-coupons to be issued to them subsequently, thereby increasing the user usage rate of e-coupons.

[0162] IV. Regarding operation S240.

[0163] In operation S240, new e-coupons are issued to target users whose number equals the number of new users. Specifically, this can be done by selecting a number of users who are predicted to be target users, and then issuing new e-coupons to these selected users.

[0164] Optionally, new e-coupons can be issued to users within the target user group whose number equals the number of new users. This method does not limit the specific method of issuing new e-coupons, nor does it limit the number of new e-coupons issued to a single target user.

[0165] Optionally, the new e-coupon can be proactively issued to the target user, or the target user can be sent a message indicating a request to issue the new e-coupon. If the target user agrees to receive the new e-coupon, the new e-coupon can be issued to the target user. For example, the target user can be sent a message to "receive e-coupon". If the target user agrees to receive the e-coupon, the new e-coupon can be issued to the target user.

[0166] In one optional embodiment, the distribution method may specifically be to recommend e-coupons to target users, or more specifically, to recommend new e-coupons. This allows new e-coupons to be issued to target users only if they agree to receive them. Since the target users are those predicted to have a high probability of using the e-coupons, the accuracy of the e-coupon recommendations can be improved.

[0167] Understandably, alternatively, e-coupons can be recommended to multiple target users until a number of target users equal to the number of new users have obtained the new e-coupon. Specifically, e-coupons can be recommended to multiple target users sequentially, and if any target user refuses to obtain the e-coupon, then the process of recommending e-coupons to a new target user can begin.

[0168] V. Regarding operation S250.

[0169] In operation S250, once it is confirmed that the new electronic coupon has been issued and the usage of the currently issued electronic coupons meets the second preset condition, the next cycle is triggered.

[0170] This method does not limit the specific content of the second preset condition. Optionally, it could be that the usage rate of currently issued e-coupons exceeds a specified usage rate threshold, thereby triggering the next cycle operation to issue additional e-coupons. Alternatively, it could be that after a specified period of time following the last issuance of e-coupons, the next cycle operation can be triggered to issue additional e-coupons; specifically, e-coupons could be issued every fixed period.

[0171] The second preset condition can be the same as the first preset condition. The explanation of the second preset condition can be found by referring to the explanation of the first preset condition.

[0172] Furthermore, the second preset condition can also differ from the first preset condition. Specifically, it could be that the percentage threshold is gradually increased as the number of loops increases, in order to increase the user usage rate. The second preset condition can be different in different loops.

[0173] Therefore, the second preset condition may optionally include at least one of the following:

[0174] 1) The percentage of users who have obtained and used electronic coupons is greater than the second percentage threshold; the second percentage threshold is positively correlated with the current cycle number.

[0175] 2) The duration between the issuance time of the new electronic coupon and the current time in this cycle is greater than the second duration threshold.

[0176] This embodiment can improve the efficiency of issuing electronic coupons and increase the user reach of electronic coupons by limiting specific second preset conditions and triggering the next cycle of issuing new electronic coupons when the usage rate or duration meets the preset conditions.

[0177] This embodiment does not limit the specific value of the second proportion threshold, nor does it limit the positive correlation between the second proportion threshold and the current loop count. For example, the second proportion threshold can increase as the current loop count increases.

[0178] This embodiment does not limit the specific method for determining the issuance time of new electronic coupons. Optionally, since multiple new electronic coupons can be issued simultaneously or not simultaneously, if the new electronic coupons are issued simultaneously, the moment when the new electronic coupons are issued simultaneously can be determined as the issuance time of the new electronic coupons. If the new electronic coupons are not issued simultaneously, the earliest, latest, or intermediate moment when the new electronic coupons are issued can be determined as the issuance time of the new electronic coupons. Furthermore, this embodiment does not limit the specific value of the second duration threshold.

[0179] Optionally, if the usage of the issued electronic coupon meets any one or more of the different second preset conditions, the loop can be triggered to execute the next iteration until the preset stop condition is met.

[0180] This method does not limit the specific way of triggering the next iteration. Optionally, it can directly trigger the re-execution of operations S220 to S250; or it can trigger the re-execution of operations S220 to S250 after waiting for a specified time. For example, the next iteration could begin the day after the current iteration ends. There may or may not be a time interval between any two adjacent iterations. The time interval between any two adjacent iterations can be used to optimize the training of the aforementioned model.

[0181] One loop in this method can be considered a stage, during which new e-coupons can be issued. Each stage can utilize user data from previous stages for optimization and training, thereby improving the accuracy of e-coupon issuance and user usage, as well as the training effect and prediction accuracy of the model, and the adaptability of the model to the e-coupon scenario.

[0182] For operation S250, this method does not limit the specific way of determining whether to trigger the next loop. Optionally, it can periodically check whether the new electronic coupons in the current loop have been issued, and whether the usage of the currently issued electronic coupons meets the second preset condition; it can also continuously monitor whether the new electronic coupons in the current loop have been issued, and whether the usage of the currently issued electronic coupons meets the second preset condition.

[0183] In any given loop, the system can remain in standby mode after issuing a new electronic coupon until the next loop is triggered. Optionally, in each loop, operations S220~S240 can be performed at the beginning of the loop, after which a check can be performed to determine whether the next loop should be triggered.

[0184] For ease of understanding, this disclosure also provides a specific application embodiment.

[0185] In one specific application embodiment, issuing vouchers (i.e., electronic vouchers in the above method embodiments) to conduct marketing activities to acquire and retain customers is a very common marketing method. After customers actively or passively obtain vouchers, they can redeem them for purchases at designated online and offline merchants within the specified time. The marketing initiator can use voucher issuance to attract and retain new customers, and at the same time, evaluate the marketing effectiveness based on voucher usage to adjust subsequent marketing strategies. Currently, there is a problem where customers receive vouchers but do not use them during the activity period, and the vouchers are only collected or invalidated after the activity ends, resulting in a waste of marketing resources and weakening the effectiveness of the marketing activity.

[0186] This embodiment can expand customer reach and improve marketing effectiveness by constructing an intelligent marketing model, given the same marketing resources and cycle. This embodiment provides an intelligent marketing method for flexibly planning marketing resources and campaign cycles. Starting from the second phase of the marketing campaign cycle, before each phase begins, the probability of unused vouchers from previous phases being used in this phase is predicted; vouchers below a certain threshold are considered unused in this phase. The number of unused vouchers in this phase is counted, and the same number of vouchers are redistributed to selected new customers in this phase, reaching a wider customer base while increasing voucher usage and enhancing marketing effectiveness.

[0187] The technical solution is explained below. (1) The marketing activity period A (unit: days) is determined by the business. (2) At the start of the activity, vouchers are randomly distributed to users who meet the activity conditions. Each user can only receive one voucher. During the marketing activity period, the voucher can be redeemed multiple times before the voucher amount is cleared. (3) The voucher redemption data is continuously monitored. The initial voucher distribution value is set to B1 (unit: vouchers), and the redemption value in the first stage is set to C1a (unit: vouchers). When the voucher usage rate in the first stage, D1 = C1a / B1, exceeds 80%, the next stage begins on the second day. The duration of the nth stage (the last stage) is set to En. E1 + E2 + E3 + ... + En-1 <A,En=A-E1-E2-E3-…-En-1。

[0188] (4) Within 2 hours of the start of the second phase, firstly, the actual consumption voucher redemption value C1b of the first phase is calculated, the full data of the first phase is collected to generate a dataset, and the consumption voucher usage probability prediction model is constructed using the random forest algorithm.

[0189] Data Acquisition: Acquire raw customer data for Phase 1, including customer information and transaction information. Customer information may include basic customer information and voucher information. Basic customer information may include: gender, nationality, age, ethnicity, region, and occupation. Voucher information may include: voucher acquisition method (active search, active claim, pop-up click, passive claim, passive acquisition, etc.), voucher acquisition stage, voucher holding duration, and voucher viewing count (whether the user actively viewed the voucher after claiming it, and the number of times they actively viewed it).

[0190] Transaction information may include: historical transaction information and the usage status of vouchers before the start of this phase. Historical transaction information may include: the number of historical transactions, the amount of each historical transaction, the total amount of historical transactions, and the type of historical transactions. The usage status of vouchers before the start of this phase may include: whether the vouchers have been used, the voucher redemption time (weekday or holiday), the amount of each transaction, the transaction type, the number of transactions (if redeemed multiple times), the total amount of transactions (if redeemed multiple times), and the voucher redemption period (which can be the time between the complete redemption date and the receipt date).

[0191] Data preprocessing: The acquired data is processed to resolve issues such as inconsistent feature specifications, information redundancy, conversion of qualitative features to quantitative features, and missing values. Detailed indicator data values ​​are generated.

[0192] A probability prediction model for the use of consumption vouchers is constructed: the model is trained using the random forest algorithm. This model corresponds to the probability prediction model for the use of electronic vouchers in the above method embodiments.

[0193] When sampling is performed using the bootstrap method, if the number of input samples is N, the original dataset will contain N samples. Data that are not selected in the sample are called out-of-bag data, which can be used to estimate the generalization error of the model.

[0194] Feature dimension M=21, preset model parameters include: feature subset constant (number of features when generating a single decision tree) m (m < M, default m = The parameters include the maximum depth of the decision tree, the number of weak learners (classification decision tree model), and the minimum number of samples per leaf node. These parameters can be fine-tuned based on the model's accuracy evaluation.

[0195] The predicted outputs of each weak learner are combined using a simple voting method, and the model is evaluated using metrics such as classification accuracy, recall, false alarm rate, and precision.

[0196] (5) Input the 21 feature values ​​of a single sample, and output the classification of the sample after passing through the model (using vouchers in this stage, not using vouchers in this stage). Based on this, predict the number of vouchers B1-C1b that were not used before the start of this stage and will not be used in this stage, F1 vouchers will be issued in this stage.

[0197] (6) Collect all data from the first stage to generate a dataset, and use the logistic regression algorithm to construct a target customer prediction model for the consumer vouchers. This model can correspond to the target user screening model in the above method embodiment.

[0198] Data Acquisition: Acquire raw customer data for Phase 1, including customer information and transaction information. Customer information may include basic customer information, such as gender, nationality, age, ethnicity, region, and occupation. Transaction information may include historical transaction data, including the number of historical transactions, the amount of each historical transaction, the total amount of historical transactions, and the type of historical transactions.

[0199] Data preprocessing: The acquired data is processed to resolve issues such as inconsistent feature specifications, information redundancy, conversion of qualitative features to quantitative features, and missing values. Detailed indicator data values ​​are generated.

[0200] To construct a prediction model for the target customers of the consumer vouchers, a logistic regression algorithm can be used. The customer data from the first stage can be divided into a training set and a test set for model training, and a prediction function can be constructed.

[0201] (7) Randomly input users who meet the activity conditions into the model, and output the results as follows: the user is a target customer of this consumption voucher, or the user is not a target customer of this consumption voucher. Reissue F1 consumption vouchers to the target customers. Then monitor the consumption voucher redemption data in the second stage. Set the consumption voucher issuance value as B2=B1+F1 (unit: vouchers), and the redemption value in the second stage as C2a (unit: vouchers). When the consumption voucher usage rate in the second stage, D2=C2a / B2, exceeds 80%, proceed to the next stage on the second day.

[0202] (8) Within 2 hours of the start of the 3rd phase, first, the actual consumption voucher redemption value C2b of the 2nd phase is counted, the full data of the first two phases is collected to generate a dataset, the consumption voucher usage probability prediction model is optimized, and the model is used to predict the number of consumption vouchers B2-C2b that were not used before the start of this phase and will not be used in this phase, F2. Then, F2 consumption vouchers are reissued in this phase.

[0203] (9) Collect all data from the first two stages to generate a dataset, optimize the prediction model for target customers of the consumer vouchers, and reissue F2 consumer vouchers to the target customers.

[0204] (10) Based on the marketing activity cycle, optimize the two models using all the data before this stage, predict the number of new consumer vouchers that can be issued in this stage, screen the target customers for consumer vouchers in this stage, so as to expand the scope of customers reached, increase the usage rate of consumer vouchers, and improve the effectiveness of marketing activities.

[0205] For ease of understanding, such as Figure 3 As shown, Figure 3 The diagram illustrates the architecture of an intelligent marketing method according to an embodiment of this disclosure. The intelligent marketing method mainly consists of a data acquisition module, a data processing module, a result output module, and an effect evaluation module. The data acquisition module acquires customer data at each stage; the data processing module primarily uses a voucher usage probability prediction model to predict whether vouchers not used before the start of the current stage will be used in this stage, and uses a voucher target customer prediction model to screen target customers for this stage; the result output module generates the number of vouchers to be reissued in this stage and distributes the vouchers to the target customers for this stage; the effect evaluation module is mainly used to perform timely statistics on the reach of customers and the voucher usage rate to evaluate the marketing effect.

[0206] For ease of understanding, such as Figure 4 As shown, Figure 4 A flowchart illustrating an intelligent marketing method according to an embodiment of the present disclosure is shown schematically. Figure 4 The document details the process for distributing vouchers for a single marketing campaign throughout its entire marketing cycle, after adopting intelligent marketing methods. Specifically, vouchers can be distributed at the start of the marketing campaign (Phase 1), and then intelligent marketing methods can be used to screen target customers for Phase 2 through Phase n and reissue vouchers until the campaign ends.

[0207] This embodiment employs intelligent marketing methods during marketing campaigns to identify unused vouchers from previous phases, predict their usage probability in the current phase, and promptly reissue vouchers predicted to be unused to selected new users, while ensuring that previously issued vouchers remain usable. Under the same marketing cycle, compared to traditional voucher distribution methods, this approach reaches more selected new target customers, avoids wasting marketing resources, increases voucher usage, and ultimately enhances marketing effectiveness.

[0208] Based on the above method embodiments, this disclosure also provides an electronic coupon issuing device. The following will be combined with... Figure 5 The device is described in detail.

[0209] Figure 5 A schematic block diagram of an electronic coupon issuing device according to an embodiment of the present disclosure is shown. Figure 5As shown, an electronic coupon issuing device 800 of this embodiment includes a trigger module 810 and a loop module 820.

[0210] The trigger module 810 is used to trigger the loop module 820 to perform an operation when it is determined that the usage of the pre-issued electronic coupons meets a first preset condition. In one embodiment, the trigger module 810 can be used to perform the operation S210 described above, which will not be repeated here.

[0211] The loop module 820 is used to repeatedly execute the following steps until a preset stopping condition is met: Based on a pre-trained electronic coupon usage probability prediction model, for the first user who has obtained an electronic coupon but has not yet used it, predict the electronic coupon usage probability; determine the number of first users whose predicted usage probability is lower than a preset probability lower limit as a baseline number; determine the number of new users who need to be issued new electronic coupons based on the determined baseline number; Based on a pre-trained target user screening model, for the new users who have not obtained electronic coupons, predict the target users; the electronic coupon usage probability of target users is higher than that of non-target users; issue new electronic coupons to the target users whose number is equal to the number of new users; and trigger the next loop execution when it is determined that the issuance of new electronic coupons has been completed and the usage of the currently issued electronic coupons meets the second preset condition. In one embodiment, the loop module 820 can be used to execute the operations S220~S250 described above, which will not be repeated here.

[0212] Optionally, the loop module 820 is specifically used to: optimize and train the current electronic coupon usage probability prediction model; based on the optimized and trained current electronic coupon usage probability prediction model, predict the electronic coupon usage probability for the first user who has obtained an electronic coupon but has not yet used it.

[0213] Optionally, the loop module 820 is specifically used to: construct a first training sample set based on the user data of users who have currently obtained electronic coupons; optimize and train the current electronic coupon usage probability prediction model based on the first training sample set; in any training sample of the first training sample set, the sample features are used to represent user information and electronic coupon information, and the sample label is used to represent that the user's electronic coupon usage probability is 0 or 1.

[0214] Optionally, the loop module 820 is specifically used to: optimize the training of the current target user screening model; and based on the optimized training of the current target user screening model, predict the target users among the new users who have not obtained electronic coupons.

[0215] Optionally, the loop module 820 is specifically used to: construct a second training sample set based on the user data of users who have obtained electronic coupons; optimize and train the current target user screening model based on the second training sample set; in any training sample of the second training sample set, the sample features are used to represent user information, and the sample labels are used to represent whether the user is a target user.

[0216] Optionally, the loop module 820 is specifically used to: determine the number of users who have obtained and used electronic coupons, and the percentage of users who have used electronic coupons among the current number of users who have obtained electronic coupons; and determine the number of new users who need to be issued new electronic coupons based on the ratio between the determined baseline number and the percentage of users who have used electronic coupons.

[0217] According to embodiments of this disclosure, any plurality of modules in trigger module 810 and loop module 820 may be combined into one module, or any one of these modules may be split into multiple modules. Alternatively, at least a portion of the functionality of one or more of these modules may be combined with at least a portion of the functionality of other modules and implemented in one module. According to embodiments of this disclosure, at least one of trigger module 810 and loop module 820 may be at least partially implemented as hardware circuitry, such as a field-programmable gate array (FPGA), a programmable logic array (PLA), a system-on-a-chip, a system-on-a-substrate, a system-on-package, an application-specific integrated circuit (ASIC), or implemented in hardware or firmware by any other reasonable means of integrating or packaging the circuitry, or implemented in any one of software, hardware, and firmware methods, or in a suitable combination of any of these. Alternatively, at least one of trigger module 810 and loop module 820 may be at least partially implemented as a computer program module, which, when run, can perform corresponding functions.

[0218] Figure 6 A block diagram schematically illustrates an electronic device suitable for implementing an electronic coupon distribution method according to an embodiment of the present disclosure.

[0219] like Figure 6 As shown, an electronic device 900 according to an embodiment of the present disclosure includes a processor 901, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 902 or a program loaded from a storage portion 908 into a random access memory (RAM) 903. The processor 901 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 901 may also include onboard memory for caching purposes. The processor 901 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present disclosure.

[0220] RAM 903 stores various programs and data required for the operation of electronic device 900. Processor 901, ROM 902, and RAM 903 are interconnected via bus 904. Processor 901 performs various operations of the method flow according to embodiments of the present disclosure by executing programs in ROM 902 and / or RAM 903. It should be noted that the programs may also be stored in one or more memories other than ROM 902 and RAM 903. Processor 901 may also perform various operations of the method flow according to embodiments of the present disclosure by executing programs stored in said one or more memories.

[0221] According to embodiments of this disclosure, the electronic device 900 may further include an input / output (I / O) interface 905, which is also connected to a bus 904. The electronic device 900 may also include one or more of the following components connected to the input / output (I / O) interface 905: an input section 906 including a keyboard, mouse, etc.; an output section 907 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 908 including a hard disk, etc.; and a communication section 909 including a network interface card such as a LAN card, modem, etc. The communication section 909 performs communication processing via a network such as the Internet. A drive 910 is also connected to the input / output (I / O) interface 905 as needed. A removable medium 911, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 910 as needed so that computer programs read from it can be installed into the storage section 908 as needed.

[0222] This disclosure also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs that, when executed, implement the method according to the embodiments of this disclosure. According to embodiments of this disclosure, the computer-readable storage medium may be a non-volatile computer-readable storage medium, such as including but not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this disclosure, the computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to embodiments of this disclosure, the computer-readable storage medium may include ROM 902 and / or RAM 903 and / or one or more memories other than ROM 902 and RAM 903 described above.

[0223] Embodiments of this disclosure also include a computer program product comprising a computer program containing program code for performing the methods shown in the flowchart. When the computer program product is run on a computer system, the program code enables the computer system to implement an electronic coupon issuance method provided by embodiments of this disclosure.

[0224] When the computer program is executed by the processor 901, it performs the functions defined in the system / apparatus of this disclosure embodiments. According to embodiments of this disclosure, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0225] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and downloaded and installed via the communication section 909, and / or installed from a removable medium 911. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0226] In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 909, and / or installed from the removable medium 911. When the computer program is executed by the processor 901, it performs the functions defined in the system of this disclosure embodiment. According to embodiments of this disclosure, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0227] According to embodiments of this disclosure, program code for executing the computer programs provided in embodiments of this disclosure can be written in any combination of one or more programming languages. Specifically, these computational programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, languages ​​such as Java, C++, Python, "C", or similar programming languages. The program code can execute entirely on a user's computing device, partially on a user's device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0228] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two connected blocks may actually be executed substantially in parallel, or they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0229] Those skilled in the art will understand that the features described in the various embodiments of this disclosure can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in this disclosure. In particular, the features described in the various embodiments of this disclosure can be combined and / or combined in various ways without departing from the spirit and teachings of this disclosure. All such combinations and / or combinations fall within the scope of this disclosure.

[0230] The embodiments of this disclosure have been described above. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of this disclosure. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of this disclosure, and all such substitutions and modifications should fall within the scope of this disclosure.

Claims

1. A method for issuing electronic coupons, characterized in that, The method includes: For pre-issued e-vouchers, if the usage of the issued e-vouchers meets the first preset condition, the following steps are executed repeatedly until the preset stop condition is met: Based on a pre-trained e-voucher usage probability prediction model, the e-voucher usage probability is predicted for the first user who has obtained an e-voucher but has not yet used it; the number of first users whose predicted usage probability is lower than a preset probability lower limit is determined as the baseline number; and based on the determined baseline number, the number of new users for whom new e-vouchers need to be issued is determined. Based on a pre-trained target user screening model, target users are predicted among new users who have not yet obtained e-coupons; the probability of e-coupon usage by target users is higher than that of non-target users. New electronic coupons will be issued to target users whose number equals the number of new users. Once it is confirmed that the issuance of new e-vouchers has been completed and the usage of currently issued e-vouchers meets the second preset condition, the next cycle will be triggered.

2. The method according to claim 1, characterized in that, The pre-trained e-coupon usage probability prediction model predicts the e-coupon usage probability for the first user who has obtained an e-coupon but has not yet used it, including: Optimize the training of the current e-coupon usage probability prediction model; based on the optimized training of the current e-coupon usage probability prediction model, predict the e-coupon usage probability for the first user who has obtained an e-coupon but has not yet used it.

3. The method according to claim 2, characterized in that, The optimization training of the current electronic coupon usage probability prediction model includes: Based on the user data of users who have already obtained electronic coupons, construct the first training sample set; Based on the first training sample set, optimize and train the current electronic coupon usage probability prediction model; In any training sample of the first training sample set, the sample features are used to characterize user information and e-coupon information, and the sample labels are used to characterize the user's e-coupon usage probability as 0 or 1.

4. The method according to claim 2 or 3, characterized in that, The electronic coupon uses a method for constructing training samples for a probabilistic prediction model, including: Among the users who have already obtained e-coupons, for any user who has obtained and used an e-coupon, a training sample is constructed based on the user data of the target user; the sample label of the constructed training sample is used to represent that the user's e-coupon usage probability is 1; Among the users who have already obtained e-coupons, for any user who has obtained but not used an e-coupon, a training sample with a pseudo-label is constructed based on the user data of the target user; the pseudo-label of the constructed training sample is used to represent that the user's probability of using the e-coupon is 0.

5. The method according to claim 1, characterized in that, The pre-trained target user screening model predicts target users among new users who have not yet received an e-coupon, including: Optimize the current target user selection model; based on the optimized current target user selection model, predict the target users among new users who have not obtained electronic coupons.

6. The method according to claim 5, characterized in that, The optimization training of the current target user selection model includes: Based on the user data of users who have already obtained electronic coupons, a second training sample set is constructed; Based on the second training sample set, optimize and train the current target user screening model; In any training sample of the second training sample set, the sample features are used to characterize user information, and the sample labels are used to characterize whether the user is the target user.

7. The method according to claim 5 or 6, characterized in that, The method for constructing the training samples of the target user screening model includes: Among the users who have already obtained e-coupons, for any user who has already obtained and used an e-coupon, a training sample is constructed based on the user data of the target user; the sample label of the constructed training sample is used to characterize that the user belongs to the target user. Among the users who have already obtained e-coupons, for any user who has obtained but not used an e-coupon, a training sample with pseudo-labels is constructed based on the user data of the target user; the pseudo-labels of the constructed training sample are used to characterize that the user belongs to a non-target user.

8. The method according to claim 1, characterized in that, The determination of the number of new users to be issued new electronic coupons based on the established baseline quantity includes: Determine the number of users who have obtained and used e-coupons, and the percentage of users who have used e-coupons out of the total number of users who have obtained e-coupons; The number of new users for whom new e-coupons need to be issued is determined based on the ratio between the established baseline number and the user usage percentage.

9. The method according to claim 1, characterized in that, The first preset condition includes at least one of the following: The percentage of users who have obtained and used e-coupons among all users who have obtained e-coupons is greater than the first percentage threshold. The duration between the issuance time of the targeted electronic coupon and the current time is greater than the first duration threshold; The second preset condition includes at least one of the following: The percentage of users who have already obtained and used e-coupons among all users who have obtained e-coupons is greater than the second percentage threshold; the second percentage threshold is positively correlated with the current cycle number. In this cycle, the duration between the issuance time of the new electronic coupon and the current time is greater than the second duration threshold.

10. An electronic coupon issuing device, characterized in that, The device includes: The trigger module is used to trigger the loop module to perform operations when the usage of the pre-issued electronic coupons meets the first preset condition. The loop module is used to repeatedly execute the following steps until a preset stopping condition is met: Based on a pre-trained e-voucher usage probability prediction model, the e-voucher usage probability is predicted for the first user who has obtained an e-voucher but has not yet used it; the number of first users whose predicted usage probability is lower than a preset probability lower limit is determined as the baseline number; and based on the determined baseline number, the number of new users for whom new e-vouchers need to be issued is determined. Based on a pre-trained target user screening model, target users are predicted among new users who have not yet obtained e-coupons; the probability of e-coupon usage by target users is higher than that of non-target users. New electronic coupons will be issued to target users whose number equals the number of new users. Once it is confirmed that the issuance of new e-vouchers has been completed and the usage of currently issued e-vouchers meets the second preset condition, the next cycle will be triggered.

11. An electronic device, comprising: One or more processors; Memory, used to store one or more computer programs. The characteristic feature is that the one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 1 to 9.

12. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 9.

13. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method according to any one of claims 1 to 9.

Images