Electronic coupon issuance method and device, readable storage medium and electronic device

Abstract

The present disclosure relates to the field of computer technology, and provides a method and device for issuing electronic coupons, a computer-readable storage medium, and an electronic device. The method includes: inputting the attribute characteristics of the electronic coupons to be issued into the offline score prediction model of the coupons to obtain the offline virtual score; obtaining the real-time collection amount and real-time usage amount of the electronic coupons to be issued from the last update time to the current time; based on the real-time collection amount and real-time usage, according to the Bayesian smoothing model, determining the real-time virtual score; determining the target virtual score according to the offline virtual score and the real-time virtual score; sorting the electronic coupons to be issued based on the target virtual score, and issuing the electronic coupons to be issued according to the sorting result. This scheme is based on the offline virtual score and the real-time virtual score, which can improve the accuracy of coupon issuance.

Description

Background Art

[0002] In e-commerce systems, providing coupons to users is a common operation method.

[0003] In the related art, coupons are mainly sorted according to their discount strength, so that coupons with large discount strength are given priority.

[0004] However, coupons with larger discounts are not necessarily the coupons that users want, so this delivery method has low accuracy.

[0005] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0006] The purpose of the present disclosure is to provide a method and device for issuing electronic coupons, a computer-readable storage medium and an electronic device, thereby improving the problem of low coupon delivery accuracy in the related art at least to a certain extent.

[0007] Other features and advantages of the present disclosure will become apparent from the following detailed description, or may be learned in part by the practice of the present disclosure.

[0008] According to a first aspect of the present disclosure, a method for issuing an electronic coupon is provided, comprising:

[0009] Inputting the attribute characteristics of the electronic coupon to be issued into the coupon offline score prediction model to obtain the offline virtual score of the electronic coupon to be issued;

[0010] Obtaining the real-time collection amount and real-time usage amount of the electronic coupon to be issued from the last update time to the current time;

[0011] Based on the real-time collection amount and the real-time usage amount, and according to the Bayesian smoothing model corresponding to the electronic coupon to be issued, determining the real-time virtual score of the electronic coupon to be issued;

[0012] Determining a target virtual score of the electronic coupon to be issued according to the offline virtual score and the real-time virtual score;

[0013] The electronic coupons to be issued are sorted based on the target virtual scores, and the electronic coupons to be issued are issued according to the sorting result.

[0014] In an exemplary embodiment of the present disclosure, based on the above solution, the step of obtaining the real-time collection amount and real-time usage amount of the electronic coupon to be issued from the last update time to the current time includes:

[0015] Obtaining the real-time collection amount and real-time usage amount of the electronic coupon to be issued from the last update time corresponding to the first issuance update cycle to the current time;

[0016] The issuing of the electronic coupons to be issued according to the sorting results includes:

[0017] In a second issuance update cycle corresponding to the current moment, issuing the electronic coupons to be issued according to the sorting result;

[0018] The current time is the update time corresponding to the second issuance update period, and the first issuance update period includes at least one of the second issuance update periods.

[0019] In an exemplary embodiment of the present disclosure, based on the above scheme, the coupon offline score prediction model is predetermined in the following manner:

[0020] Acquire the attribute characteristics of the electronic coupons that have been issued and the conversion rate of the electronic coupons that have been issued;

[0021] Using the attribute features of the electronic coupons that have been issued as input samples and the conversion rate of the electronic coupons that have been issued as optimization target values, the gradient boosting iterative decision tree model is trained to obtain the offline prediction model of the coupons;

[0022] The conversion rate of the electronic coupon that has been issued is determined according to the ratio between the historical usage amount and the historical collection amount of the electronic coupon that has been issued.

[0023] In an exemplary embodiment of the present disclosure, based on the above solution, the Bayesian smoothing model corresponding to the electronic coupon to be issued is determined by the following formula:

[0024]

[0025] Wherein, U is the real-time usage of the electronic coupon to be issued from the last update time corresponding to the first issuance update cycle to the current time, G is the real-time collection amount of the electronic coupon to be issued from the last update time corresponding to the first issuance update cycle to the current time, and β are smoothing parameters in the Bayesian smoothing model;

[0026] The smoothing parameter in the Bayesian smoothing model corresponding to each of the electronic coupons to be issued is determined in the following manner:

[0027] Acquire historical real-time data of the electronic coupon to be issued, the historical real-time data including the collection amount and usage amount of the electronic coupon to be issued between the update time of each first issuance cycle and the update time corresponding to each second issuance update cycle within the first issuance cycle within a recent preset time;

[0028] The historical real-time data is used as prior data, and based on moment estimation and maximum expectation algorithm, the smoothing parameters in the Bayesian smoothing model corresponding to the electronic coupon to be issued are obtained.

[0029] In an exemplary embodiment of the present disclosure, based on the above solution, the method further includes:

[0030] When the electronic coupon to be issued does not have the historical real-time data, determining that the electronic coupon to be issued is a new coupon;

[0031] The median or average of the smoothing parameters corresponding to other electronic coupons to be issued that have the historical real-time data is determined as the smoothing parameter in the Bayesian smoothing model corresponding to the new coupon.

[0032] In an exemplary embodiment of the present disclosure, based on the above solution, the step of sorting the to-be-issued electronic coupons based on the target virtual scores includes:

[0033] Determine the collection amount and usage amount of each electronic coupon to be issued from the last update time corresponding to the first issuance update cycle to the current time;

[0034] When the claimed amount is greater than a first preset threshold and the used amount is less than a second preset threshold, determining that the electronic coupon to be issued is a tail coupon; otherwise, determining that the electronic coupon to be issued is a normal coupon;

[0035] Based on the target virtual score, the normal coupons and the tail coupons are sorted in descending order to determine the sorting result of the electronic coupons to be issued;

[0036] Among them, in the sorting result, the sorting order of the normal ticket with the smallest target virtual score is placed before the sorting order of the tail ticket with the largest target virtual score.

[0037] In an exemplary embodiment of the present disclosure, based on the above solution, the first preset threshold and the second preset threshold are determined in the following manner:

[0038] Obtain historical real-time data of each electronic coupon to be issued to generate a training sample set;

[0039] Performing unsupervised learning training on the clustering model according to the training sample set to perform binary classification on the electronic coupons to be issued;

[0040] Determining the first preset threshold and the second preset threshold according to the classification result;

[0041] The historical real-time data includes the collection amount and usage amount of the electronic coupons to be issued between the update time of each first issuance cycle and the update time corresponding to each second issuance update cycle within the first issuance cycle within the most recent preset time.

[0042] In an exemplary embodiment of the present disclosure, based on the above solution, the method further includes:

[0043] In a target time period between an update time of each of the first issuance update cycles and an update time of a first second issuance update cycle within the first issuance update cycle, the electronic coupons to be issued are sorted in descending order based on the offline virtual scores, so that the electronic coupons to be issued are issued within the target time period according to the sorting result.

[0044] In an exemplary embodiment of the present disclosure, based on the above solution, determining the target virtual score of the electronic coupon to be issued according to the offline virtual score and the real-time virtual score includes:

[0045] Determining a first product between the offline virtual score and a first preset weight;

[0046] Determine a second product between the real-time virtual score and a second preset weight;

[0047] The target virtual score of the electronic coupon to be issued is determined according to the sum of the first product and the second product.

[0048] In an exemplary embodiment of the present disclosure, based on the above solution, issuing the to-be-issued electronic coupons according to the ranking results includes:

[0049] According to the descending sorting result of the target virtual scores, the first K electronic coupons to be issued are selected from the electronic coupons to be issued for issuance, wherein K is a positive integer.

[0050] In an exemplary embodiment of the present disclosure, based on the aforementioned scheme, the attribute characteristics include one or more of the category, denomination, limit, discount strength, popularity value of the covered inventory unit, price of the covered inventory unit, covered commodity category, covered commodity brand, and covered store of the electronic coupon.

[0051] According to a second aspect of the present disclosure, there is provided an electronic coupon issuing device, comprising:

[0052] An offline virtual score determination module is configured to input the attribute characteristics of the electronic coupon to be issued into the coupon offline score prediction model to obtain the offline virtual score of the electronic coupon to be issued;

[0053] A real-time data acquisition module is configured to acquire the real-time collection amount and real-time usage amount of the electronic coupon to be issued from the last update time to the current time;

[0054] A real-time virtual score determination module is configured to determine the real-time virtual score of the electronic coupon to be issued based on the real-time collection amount and the real-time usage amount and according to the Bayesian smoothing model corresponding to the electronic coupon to be issued;

[0055] a target virtual score determination module, configured to determine the target virtual score of the electronic coupon to be issued according to the offline virtual score and the real-time virtual score;

[0056] The coupon issuing module is configured to sort the electronic coupons to be issued based on the target virtual score, and issue the electronic coupons to be issued according to the sorting result.

[0057] According to a third aspect of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored. When the program is executed by a processor, the method for issuing electronic coupons as described in the first aspect of the above embodiment is implemented.

[0058] According to a fourth aspect of an embodiment of the present disclosure, an electronic device is provided, comprising: a processor; and a storage device for storing one or more programs, which, when executed by the one or more processors, enables the one or more processors to implement the method for issuing electronic coupons as described in the first aspect of the above embodiment.

[0059] It can be seen from the above technical solutions that the electronic coupon issuance method, the electronic coupon issuance device, and the computer-readable storage medium and electronic device for implementing the electronic coupon issuance method in the exemplary embodiment of the present disclosure have at least the following advantages and positive effects:

[0060] In the technical solutions provided by some embodiments of the present disclosure, first, the attribute characteristics of the electronic coupon to be issued are input into the offline score prediction model of the coupon to obtain the offline virtual score of the electronic coupon to be issued, then the real-time collection amount and real-time usage amount of the electronic coupon to be issued from the last update time to the current time are obtained, and based on the real-time collection amount and the real-time usage amount, according to the Bayesian smoothing model corresponding to the electronic coupon to be issued, the real-time virtual score of the electronic coupon to be issued is determined, and the target virtual score of the electronic coupon to be issued is determined according to the offline virtual score and the real-time virtual score, and finally, the electronic coupon to be issued is sorted based on the target virtual score, and the electronic coupon to be issued is issued according to the sorting result. Compared with the related art, the present disclosure can improve the accuracy of coupon issuance based on the attribute characteristics of the coupon and the real-time usage data of the coupon.

[0061] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] The accompanying drawings herein are incorporated into the specification and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the specification are used to explain the principles of the present disclosure. Obviously, the accompanying drawings described below are only some embodiments of the present disclosure, and for ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without creative work.

[0063] Figure 1 A schematic diagram showing a flow chart of an electronic coupon issuing method in an exemplary embodiment of the present disclosure;

[0064] Figure 2 A flowchart showing a method for training a coupon offline score prediction model in an exemplary embodiment of the present disclosure is shown;

[0065] Figure 3 A framework diagram showing the training of an offline coupon prediction model in an exemplary embodiment of the present disclosure;

[0066] Figure 4 A flowchart showing a method for determining a smoothing parameter in a Bayesian smoothing model corresponding to each coupon to be issued in an exemplary embodiment of the present disclosure;

[0067] Figure 5 A schematic flow chart showing a method for determining a target virtual score in an exemplary embodiment of the present disclosure;

[0068] Figure 6A flowchart showing a method for sorting electronic coupons to be issued in an exemplary embodiment of the present disclosure is shown;

[0069] Figure 7 A schematic flow chart showing a method for determining a first preset threshold and a second preset threshold in an exemplary embodiment of the present disclosure;

[0070] Figure 8 A flowchart showing another method for sorting coupons in an exemplary embodiment of the present disclosure is shown;

[0071] Fig. 9 A schematic diagram showing the structure of an electronic coupon issuing device in an exemplary embodiment of the present disclosure;

[0072] Fig.10 A schematic diagram showing the structure of a computer storage medium in an exemplary embodiment of the present disclosure;

[0073] Fig.11 A schematic structural diagram of an electronic device in an exemplary embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0074] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as being limited to the examples set forth herein; on the contrary, these embodiments are provided so that the present disclosure will be more comprehensive and complete, and the concepts of the example embodiments are fully conveyed to those skilled in the art. The described features, structures, or characteristics may be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present disclosure. However, those skilled in the art will appreciate that the technical solutions of the present disclosure may be practiced while omitting one or more of the specific details, or other methods, components, devices, steps, etc. may be adopted. In other cases, known technical solutions are not shown or described in detail to avoid obscuring various aspects of the present disclosure.

[0075] The terms "a", "an", "the" and "said" are used in this specification to indicate the presence of one or more elements / components / etc.; the terms "including" and "having" are used to express an open-ended inclusion and mean that additional elements / components / etc. may exist in addition to the listed elements / components / etc.; the terms "first" and "second" etc. are used only as labels and are not intended to limit the quantity of their objects.

[0076] In addition, the accompanying drawings are only schematic illustrations of the present disclosure and are not necessarily drawn to scale. The same reference numerals in the drawings represent the same or similar parts, and their repeated descriptions will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities.

[0077] In e-commerce systems, issuing coupons to users is a common operation method. When issuing coupons, it is usually divided into two stages: recall and sorting. In the recall stage, all coupons in the coupon pool are initially screened according to user preferences to narrow the range of coupons that can be issued to users; in the sorting stage, the recalled coupons are sorted, and coupons are issued according to the sorting results, that is, the best coupons are recommended to users.

[0078] In the related art, the sorting stage is to select the best coupons from the recalled coupons. The most common way is to sort them according to the discount strength of the coupons, and finally select the coupons with the highest discount strength for distribution.

[0079] However, since coupons with large discounts are not necessarily the coupons that users need, this method of distributing coupons has low accuracy.

[0080] In the embodiments of the present disclosure, a method for issuing electronic coupons is first provided, which overcomes the defects existing in the above-mentioned related technologies at least to a certain extent.

[0081] Figure 1 A flow chart showing a method for issuing electronic coupons in an exemplary embodiment of the present disclosure is shown. Figure 1 , the method comprising:

[0082] Step S110, inputting the attribute characteristics of the electronic coupon to be issued into the coupon offline score prediction model to obtain the offline virtual score of the electronic coupon to be issued;

[0083] Step S120, obtaining the real-time collection amount and real-time usage amount of the electronic coupon to be issued from the last update time to the current time;

[0084] Step S130, based on the real-time collection amount and the real-time usage amount, and according to the Bayesian smoothing model corresponding to the electronic coupon to be issued, determining the real-time virtual score of the electronic coupon to be issued;

[0085] Step S140, determining the target virtual score of the electronic coupon to be issued according to the offline virtual score and the real-time virtual score;

[0086] Step S150: sorting the electronic coupons to be issued based on the target virtual scores, and issuing the electronic coupons to be issued according to the sorting result.

[0087] exist Figure 1 In the technical solution provided by the illustrated embodiment, first, the attribute characteristics of the electronic coupon to be issued are input into the offline score prediction model of the coupon to obtain the offline virtual score of the electronic coupon to be issued, then the real-time collection amount and real-time usage amount of the electronic coupon to be issued from the last update time to the current time are obtained, and based on the real-time collection amount and the real-time usage amount, according to the Bayesian smoothing model corresponding to the electronic coupon to be issued, the real-time virtual score of the electronic coupon to be issued is determined, and the target virtual score of the electronic coupon to be issued is determined according to the offline virtual score and the real-time virtual score, and finally, the electronic coupon to be issued is sorted based on the target virtual score, and the electronic coupon to be issued is issued according to the sorting result. Compared with the related art, the present disclosure can improve the accuracy of coupon issuance based on the attribute characteristics of the coupon and the real-time usage data of the coupon.

[0088] The following Figure 1 The specific implementation methods of each step in the embodiment shown are described in detail:

[0089] In step S110, the attribute features of the electronic coupon to be issued are input into the coupon offline score prediction model to obtain the offline virtual score of the electronic coupon to be issued.

[0090] In an exemplary embodiment, the attribute characteristics of the electronic coupon to be issued include one or more of the category, denomination, limit, discount strength, popularity value of the covered inventory unit, price of the covered inventory unit, covered commodity category, covered commodity brand, and covered store of the electronic coupon.

[0091] Among them, the category of the electronic coupon can be understood as the category of the electronic coupon itself, such as the full-discount coupon category, cash coupon category, etc. The limit of the electronic coupon can be understood as the minimum consumption amount for using the coupon. For example, if the limit is 99 yuan, it means that when the consumption amount is 99 yuan or more, the electronic coupon can be used to offset part of the payment amount, otherwise the electronic coupon cannot be used.

[0092] For example, the coupon offline score prediction model may be obtained by pre-training through the attribute features of historical coupons that have been issued. Figure 2 A flow chart showing a method for training a coupon offline score prediction model in an exemplary embodiment of the present disclosure is shown. Figure 2 , the method may include steps S210 to S220. Wherein:

[0093] In step S210, the attribute characteristics of the electronic coupons that have been issued and the conversion rate of the electronic coupons that have been issued are obtained.

[0094] For example, each electronic coupon may correspond to a start date and an end date, wherein the start date indicates the date corresponding to the first day on which the electronic coupon can be issued, and the end date indicates the date corresponding to the last day on which the electronic coupon can be issued, that is, the electronic coupon will no longer be issued after the last day.

[0095] In an optional implementation, the electronic coupons whose issuance has ended include all electronic coupons whose issuance end date is less than the training date of the coupon offline score prediction model.

[0096] For example, the offline score prediction model of coupons can be obtained by training with the attribute features of the electronic coupons that have been issued before the day, so as to update the offline score prediction model of coupons on a daily basis. Then, the offline score prediction model of coupons updated on the day is used to predict the offline virtual scores of the electronic coupons to be issued tomorrow.

[0097] Of course, the coupon offline score prediction model may also be updated in other time units, such as in "weeks", "months" or any other fixed period, or the coupon offline score prediction model may be updated irregularly at any time, and this exemplary embodiment does not specifically limit this.

[0098] In another optional implementation, the electronic coupons that have ended issuance may include electronic coupons that ended issuance within a preset historical time interval closest to the update date of the coupon offline score prediction model, such as electronic coupons that ended issuance within 60 days before the update date.

[0099] The attribute characteristics of the electronic coupons that have been issued are the same as the attribute characteristics of the electronic coupons to be issued, and will not be described in detail here.

[0100] In an exemplary embodiment, the conversion rate of the electronic coupon that has been issued is determined according to the ratio between the historical usage and the historical collection of the electronic coupon that has been issued, that is, the ratio between the historical total usage and the historical total collection of the electronic coupon that has been issued is its corresponding conversion rate. For example, a coupon is issued from July 1, 2021 to July 15, 2021. The conversion rate of the coupon is the ratio between the usage and the collection within 15 days from July 1, 2021 to July 15, 2021.

[0101] Next, in step S220, the attribute features of the electronic coupons that have been issued are used as input samples, and the conversion rate of the electronic coupons that have been issued is used as the optimization target value, and the gradient boosting iterative decision tree model is trained to obtain the coupon offline score prediction model.

[0102] The gradient boosting iterative decision tree model can be understood as a GBDT (Gradient Boosting Decision Tree) model. The attribute features of each electronic coupon that has been issued can be used as input, and the conversion rate of each electronic coupon that has been issued can be used as the optimization target value of the model. The established GDBT regression model is trained to obtain a coupon offline score prediction model.

[0103] Before using the attribute features and conversion rates of the electronic coupons that have been issued to train the constructed GDBT model, the attribute features of the electronic coupons that have been issued can also be preprocessed and the preprocessed attribute features can be used for training.

[0104] For example, the table storing electronic coupons can be associated with the table storing the popularity of the inventory units of the goods covered by the electronic coupons and the prices of the goods covered by the electronic coupons, thereby generating cross-features of the inventory units of the goods covered by the electronic coupons. The category of goods covered by the electronic coupons, the brand of the covered goods, and the covered store features can also be reduced in dimension by using an embedding method, so that these features have mutual correlation in the reduced-dimensional space, thereby improving the prediction accuracy of the model. Among them, the method of reducing the dimension of features by using an embedding method can refer to the existing technology and will not be repeated here.

[0105] After preprocessing the features, the preprocessed attribute features, such as the category, denomination, limit, discount strength, and cross-features of the inventory units of the covered goods of the electronic coupons, as well as the attribute features after dimensionality reduction processing of the category, brand, and store features of the covered goods of the electronic coupons and the conversion rate corresponding to the electronic coupons can be used as sample data to train a coupon offline prediction model.

[0106] For example, Figure 3 FIG. 1 is a framework diagram showing the training of the coupon offline score prediction model in an exemplary embodiment of the present disclosure. Figure 3 ,The framework mainly includes three parts, namely model input features 31, model optimization objectives 32, and model output 33.

[0107] Among them, the model input features 31 include two categories: coupon metadata and coupon restriction sku (Stock Keeping Unit, i.e. the above-mentioned inventory unit) data. Specifically, the coupon metadata may include the face value, limit, coupon type, third-level categories covered by the coupon, covered stores, etc., and the coupon restriction sku data may include the cross-features obtained by associating the tables of available sku, dynamic sales sku, order-matching sku, popularity value, discount strength, and prices of covered goods.

[0108] The model optimization target 32 ​​is the conversion rate corresponding to each coupon in the training sample, and the model output 33 is the offline virtual score of the coupon. Using the model input feature 31 as the input sample and the model optimization target 32 ​​as the input sample label, the GBDT model is trained in supervised learning to obtain the coupon offline score prediction model.

[0109] Through the above steps S210 to S220, the coupon offline score prediction model can be obtained by training the attribute features of the coupon. Since the attribute features of the coupon can include many aspects of the coupon, not just the discount strength features, it can improve the accuracy of the coupon offline virtual score prediction.

[0110] After the coupon offline score prediction model is trained, the offline virtual score of the electronic coupon to be issued can be obtained based on the coupon offline score prediction model. That is, after the attribute characteristics of each electronic coupon to be issued are input into the coupon offline score prediction model, the output value of the model is the offline virtual score of the electronic coupon to be issued.

[0111] Continue to refer Figure 1 In step S120, the real-time collection amount and real-time usage amount of the electronic coupon to be issued from the last update time to the current time are obtained.

[0112] In an exemplary implementation, the target virtual score of the electronic coupons to be issued may be updated periodically to select the electronic coupons that need to be issued currently from all the electronic coupons to be issued.

[0113] In the present disclosure, two issuance update cycles may be included, namely, a first issuance update cycle and a second issuance update cycle. The first issuance update cycle includes at least one second issuance update cycle. In other words, the time interval between each update moment of the second issuance update cycle is less than or equal to the time interval between each update moment of the first issuance update cycle. The target virtual score of the electronic coupon to be issued may be updated at the update moments corresponding to the two issuance update cycles.

[0114] For example, the first issuance update cycle is 8 hours apart, that is, 0:00, 8:00, and 16:00 every day are the update times corresponding to the first issuance update cycle, and the second issuance update cycle is 1 hour apart, that is, each hour in each first issuance update cycle is the update time corresponding to the second issuance update cycle, such as 1:00, 2:00, 3:00, 4:00, 5:00, 6:00, and 7:00 are the update times corresponding to the second issuance update cycle included in the first issuance update cycle from 0:00 to 8:00.

[0115] Exemplarily, the specific implementation of step S120 may include: obtaining the real-time collection amount and real-time usage amount of the electronic coupon to be issued from the last update time corresponding to the first issuance update cycle to the current time. The current time includes the update time corresponding to the second issuance update cycle within the current first issuance update cycle.

[0116] Taking the example that the last update time corresponding to the first issuance update cycle is 8:00 on July 11, 2021, the current time may be 9:00, 10:00, 11:00, 12:00, 13:00, 14:00, and 15:00 on July 11, 2021. Continuing to take the current time as 9:00 and 10:00 as an example, when the current time is 9:00 on July 11, 2021, the collection amount and usage amount of each electronic coupon to be issued between 8:00 and 9:00 can be obtained in step S120, so as to determine which electronic coupons can be issued between 9:00 and 10:00. When the current time is 10:00 on July 11, 2021, the collection amount and usage amount of each electronic coupon to be issued between 8:00 and 10:00 can be obtained in step S120, so as to determine which electronic coupons can be issued between 10:00 and 11:00.

[0117] It should be noted that only one update cycle may be included, that is, the first issuance update cycle and the second issuance update cycle may be the same, and this exemplary embodiment does not specifically limit this.

[0118] Continue to refer Figure 1 Next, in step S130, based on the real-time collection amount and the real-time usage amount, according to the Bayesian smoothing model corresponding to the electronic coupon to be issued, the real-time virtual score of the electronic coupon to be issued is determined.

[0119] In an exemplary embodiment, the Bayesian smoothing model corresponding to each electronic coupon to be issued can be determined by the following formula:

[0120]

[0121] Wherein, U is the real-time usage of the electronic coupon to be issued from the last update time corresponding to the first issuance update cycle to the current time, G is the real-time collection amount of the electronic coupon to be issued from the last update time corresponding to the first issuance update cycle to the current time, and β are smoothing parameters in the Bayesian smoothing model corresponding to the electronic coupon to be issued.

[0122] Figure 4 A flow chart showing a method for determining a smoothing parameter in a Bayesian smoothing model corresponding to each coupon to be issued in an exemplary embodiment of the present disclosure. Figure 4 , the method may include steps S410 to S420. Wherein:

[0123] In step S410, historical real-time data of the electronic coupons to be issued is obtained.

[0124] In an exemplary embodiment, the historical real-time data of the electronic coupons to be issued includes the collection amount and usage amount of the electronic coupons to be issued between the update time of each first issuance cycle and the update time corresponding to each second issuance update cycle within the first issuance cycle within the most recent preset time.

[0125] Among them, the most recent preset time can be customized according to needs, such as obtaining historical real-time data of the electronic coupons to be issued within the past 7 days.

[0126] Taking the update time corresponding to the first issuance update cycle as 0:00, 8:00 and 16:00 every day, and the second issuance update cycle as an example of an interval of 1 hour in each first issuance update cycle, the collection amount and usage amount in each time period corresponding to 0:00 to 1:00, 0:00 to 2:00, 0:00 to 3:00, 0:00 to 4:00, 0:00 to 5:00, 0:00 to 6:00, 0:00 to 7:00, 8:00 to 9:00, 8:00 to 10:00, 8:00 to 11:00, 8:00 to 12:00, 8:00 to 13:00, 8:00 to 14:00, 8:00 to 15:00, 16:00 to 17:00, 16:00 to 18:00, 16:00 to 19:00, 16:00 to 20:00, 16:00 to 21:00, 16:00 to 22:00, and 16:00 to 23:00 of the electronic coupon to be issued in the past 7 days can be obtained respectively, and the amount is used as the historical real-time data of the electronic coupon to be issued to form a historical real-time data set.

[0127] When the first distribution update cycle and the second distribution update cycle are the same, the historical real-time data can be understood as the collection amount and usage amount of the electronic coupons to be distributed between each two adjacent update moments within the recent preset time.

[0128] After obtaining the historical real-time data of the electronic coupons to be issued, in step S420, the historical real-time data is used as prior data, and based on moment estimation and maximum expectation algorithm, the smoothing parameters in the Bayesian smoothing model corresponding to the electronic coupons to be issued are obtained.

[0129] For example, the historical real-time data in the above historical real-time data set can be used as prior data, and based on the idea of ​​moment estimation, the smoothing parameters in the Bayesian smoothing model corresponding to the electronic coupon to be issued can be obtained. The initial value of and β is then used as the initial value in the maximum expectation algorithm. Then, the historical real-time data is used to obtain the smoothing parameters in the Bayesian smoothing model based on the idea of ​​the maximum expectation algorithm. and β.

[0130] For example, the amount of collection and usage between the update time of each first distribution cycle and the update time of each second distribution update cycle in the first distribution cycle in the past 7 days is taken as a priori data. The collection amount corresponding to each priori data is recorded as G, and the corresponding usage amount is recorded as U. Then the conversion rate corresponding to the priori data is R = U / G. The mean of R is recorded as mean, and the variance of R is recorded as var. Then, according to the moment estimation, the initialization is obtained. and β are: The initialized and β as initial values, and are introduced into the EM (Expectation-Maximum) algorithm to update according to the following formulas (2) and (3) respectively: and β until convergence, the final and β, the and β are the smoothing parameters in the Bayesian smoothing model corresponding to the electronic coupon to be issued. The update formula of the EM algorithm is:

[0131]

[0132]

[0133] In formula (2) and formula (3), Indicates the time since the last update Value, β old Represents the β value after the last update. G i represents the amount of collection in the i-th prior data, U i represents the usage in the i-th prior data, and ψ(g) represents the Digamma function.

[0134] In an exemplary embodiment, when the electronic coupon to be issued does not have the historical real-time data, the electronic coupon to be issued is determined to be a new coupon; and the median or average of the smoothing parameters corresponding to other electronic coupons to be issued that have the historical real-time data is determined as the smoothing parameter in the Bayesian smoothing model corresponding to the new coupon.

[0135] For example, for some coupons to be issued, they may be new coupons issued on the same day, so there is no historical real-time data in the past 7 days. Then the median or average of the smoothing parameters corresponding to other electronic coupons to be issued that have historical real-time data in the past 7 days can be used as the smoothing parameters in the Bayesian smoothing model corresponding to the new coupons. Of course, other statistical parameters of the smoothing parameters corresponding to other electronic coupons to be issued can also be used as the smoothing parameters corresponding to the new coupons, and this exemplary embodiment does not make special restrictions on this.

[0136] In the operation of coupons in e-commerce, the conversion rate indicator is effective under large amounts of data. Ideally, if the number of coupons collected is 10,000 and the number of coupons used is 100, then the conversion rate of the coupon is 1%, which is effective. However, in real-time streaming, there may be situations where the number of coupons collected and used is very small. For example, the number of coupons collected is 5 and the number of coupons used is 2. The calculated coupon conversion rate is 40%, but at this time, this data is mathematically invalid and does not conform to the "law of large numbers" because in the "law of large numbers", under the condition that the experiment remains unchanged, the experiment is repeated many times, and the frequency of random events is close to its probability. However, the latter only collected 5 coupons, which does not meet the condition of "repeated experiments many times".

[0137] In the present disclosure, the Bayesian smoothing model can be used to reasonably process data that does not conform to the law of large numbers. In other words, for a certain coupon, even if its real-time collection and usage do not conform to the "law of large numbers", the collection and usage can be processed through the Bayesian smoothing model, so that a reasonable real-time score can be obtained.

[0138] Exemplarily, through the above steps S410 to S420, the Bayesian smoothing model corresponding to each electronic coupon to be issued can be determined. For each electronic coupon to be issued, the real-time usage and real-time collection of the electronic coupon to be issued can be brought into its corresponding Bayesian smoothing model to obtain the real-time virtual score of the electronic coupon to be issued.

[0139] Next, in step S140, the target virtual score of the electronic coupon to be issued is determined according to the offline virtual score and the real-time virtual score.

[0140] For example, Figure 5A flow chart showing a method for determining a target virtual score in an exemplary embodiment of the present disclosure is shown. Figure 5 , the method may include steps S510 to S530. In step S510, a first product between the offline virtual score and a first preset weight is determined; in step S520, a second product between the real-time virtual score and a second preset weight is determined; in step S530, a target virtual score of the electronic coupon to be issued is determined according to the sum of the first product and the second product.

[0141] For example, a first preset weight A corresponding to the offline virtual score and a second preset weight B corresponding to the real-time virtual score may be preset. Then, the offline virtual score and the real-time virtual score corresponding to each coupon to be issued are merged by the following formula (4) to obtain the target virtual score.

[0142] allscore=A*offlinescore+B*realscore (4)

[0143] In formula (4), offlinescore represents the offline virtual score, realscore represents the real-time virtual score, and allscore represents the target virtual score. The value of the first preset weight A and the value of the second preset weight B can be customized according to needs, and this exemplary embodiment does not make any special restrictions on this.

[0144] In step S150, the electronic coupons to be issued are sorted based on the target virtual scores, and the electronic coupons to be issued are issued according to the sorting result.

[0145] In the present disclosure, based on the historical real-time data of electronic coupons, a clustering model in an unsupervised learning algorithm can be constructed, so that the tail coupons in the electronic coupons to be issued can be determined based on the clustering model to downgrade the tail coupons for sorting the electronic coupons to be issued in step S150.

[0146] Next, combine Figure 6 and Figure 7 The specific implementation of the above-mentioned step S510 is described.

[0147] For example, Figure 6 A flow chart showing a method for sorting electronic coupons to be issued in an exemplary embodiment of the present disclosure is shown, with reference to Figure 6 , the method may include steps S610 to S630. Wherein:

[0148] In step S610, the collection amount and usage amount of each electronic coupon to be issued from the last update time corresponding to the first issuance update cycle to the current time are determined.

[0149] Exemplarily, in step S610, for each electronic coupon to be issued, its collection amount and usage amount from the last update time corresponding to the first issuance update cycle to the current time are the same as those in the above step S120, and will not be repeated here.

[0150] In step S620, when the claimed amount is greater than a first preset threshold and the used amount is less than a second preset threshold, the electronic coupon to be issued is determined to be a tail coupon; otherwise, the electronic coupon to be issued is determined to be a normal coupon.

[0151] In an exemplary embodiment, the tail coupons can be understood as pending electronic coupons that need to be downgraded in the subsequent sorting stage. The normal coupons can be understood as electronic coupons that can be directly sorted according to the target virtual score in the subsequent sorting stage.

[0152] Next, combine Figure 7 The method for determining the first preset threshold and the second preset threshold in step S620 is described below. Figure 7 , the method for determining the first preset threshold and the second preset threshold may include steps S710 to S730. Among them:

[0153] In step S710, historical real-time data of each electronic coupon to be issued is obtained to generate a training sample set.

[0154] The historical real-time data includes the collection amount and usage amount of the electronic coupons to be issued between the update time of each first issuance cycle and the update time corresponding to each second issuance update cycle within the first issuance cycle within the most recent preset time.

[0155] In other words, the historical real-time data of each electronic coupon to be issued in step S710 is the same as the historical real-time data in the above step S410, except that step S410 is for each electronic coupon to be issued separately, while the historical real-time data of all electronic coupons to be issued is obtained in step S710.

[0156] For example, the historical real-time data of each electronic coupon to be issued obtained in the above step S410 may be aggregated to generate a set, and the set is used as a training sample set.

[0157] In step S720, unsupervised learning training is performed on the clustering model according to the training sample set to perform binary classification on the electronic coupons to be issued.

[0158] In an exemplary embodiment, the clustering model may include a k-means clustering algorithm, a DBSCAN (Density—Based Spatial Clustering of Application with Noise) algorithm, or the like.

[0159] Taking the k-means clustering algorithm as an example, since the electronic coupons to be issued need to be divided into normal coupons and tail coupons in the present disclosure, the K value in the k-means algorithm is taken as 2, and then the unsupervised learning training is performed using the training sample set based on the process of the k-means clustering method in the prior art. The k-means clustering process can refer to the prior art and will not be repeated here.

[0160] In step S730, the first preset threshold and the second preset threshold are determined according to the classification result.

[0161] Exemplarily, after the k-means clustering method performs binary classification on the electronic coupons to be issued, there will be a dividing line between the two types of electronic coupons, and the dividing line corresponds to a value of the amount received and a value of the amount used. The value of the amount received corresponding to the dividing line is used as the first preset threshold, and the value of the amount used corresponding to the dividing line is used as the second preset threshold.

[0162] Through the above steps S710 to S730, the first preset threshold and the second preset threshold can be determined according to the historical real-time data of the electronic coupons to be issued. Then in step S620, the category label of each electronic coupon to be issued can be determined based on the first preset threshold and the second preset threshold, that is, whether the electronic coupon to be issued is a tail coupon or a normal coupon.

[0163] Continue to refer Figure 6 In step S630, based on the target virtual score, the normal coupons and the tail coupons are sorted in descending order to determine the sorting result of the electronic coupons to be issued;

[0164] Among them, in the sorting result, the sorting order of the normal ticket with the smallest target virtual score is placed before the sorting order of the tail ticket with the largest target virtual score.

[0165] For example, after determining the tail coupons and normal coupons, all the tail coupons can be sorted in descending order based on the target virtual score within all the tail coupons, and all the normal coupons can be sorted in descending order based on the target virtual score within all the normal coupons, and then all the normal coupons are placed before all the tail coupons, thereby achieving downgrading of the tail coupons.

[0166] Through the above steps S610 to S630, the coupons can be sorted and optimized according to the real-time data of the coupons, thereby further improving the accuracy of coupon issuance or recommendation.

[0167] After the ranking result of the electronic coupons to be issued is determined, the electronic coupons to be issued may be issued according to the ranking result.

[0168] Exemplarily, issuing the electronic coupons to be issued according to the sorting result includes: selecting first K electronic coupons to be issued from the electronic coupons to be issued for issuance according to the descending sorting result of the target virtual scores, where K is a positive integer.

[0169] For example, the first K electronic coupons to be issued can be selected from the electronic coupons to be issued according to the sorting result after the tail coupons are degraded. The K value can be customized according to user needs, and this exemplary embodiment does not make any special restrictions on this.

[0170] For example, Figure 8 A flowchart showing another method for sorting coupons in an exemplary embodiment of the present disclosure is shown. Figure 8 , the method may include steps S810 to S870.

[0171] In step S810, the coupon offline points are determined;

[0172] In step S820, the real-time points of the coupon are determined;

[0173] In step S830, the offline score of the coupon and the real-time score of the coupon are merged to determine the merged score of the coupon;

[0174] In step S840, real-time coupon data is obtained;

[0175] In step S850, it is determined whether the coupon is a tail coupon according to the real-time coupon data. If yes, the process goes to step S860, otherwise, the process goes to step S870;

[0176] In step S860, the coupons are sorted in descending order;

[0177] In step S870, the coupons are sorted normally.

[0178] In an exemplary embodiment, Figure 8The offline score in the above can be understood as the offline virtual score, the real-time score can be understood as the real-time virtual score, and the coupon real-time data can include the real-time collection amount and real-time usage amount of the electronic coupon to be issued in step S220 from the last update time to the current time, which will not be described here. The fusion score can be understood as the target virtual score.

[0179] In an exemplary implementation, issuing the electronic coupons to be issued according to the sorting result in step S150 includes: issuing the electronic coupons to be issued according to the sorting result within a second issuance update cycle corresponding to the current moment; wherein the current moment is an update moment corresponding to the second issuance update cycle.

[0180] For example, if the current time is 1 o'clock, the electronic coupons to be issued can be issued between 1 o'clock and 2 o'clock according to the sorting result, and if the current time is 2 o'clock, the electronic coupons to be issued can be issued between 2 o'clock and 3 o'clock according to the sorting result. That is, between two adjacent update times corresponding to the second issuance update cycle, the electronic coupons to be issued can be issued according to the sorting result of the target virtual score.

[0181] In another exemplary embodiment, within a target time period between an update time of each of the first issuance update cycles and an update time of a first second issuance update cycle within the first issuance update cycle, the electronic coupons to be issued are sorted in descending order based on the offline virtual scores, so that the electronic coupons to be issued are issued within the target time period according to the sorting result.

[0182] Continuing with the example of every 8 hours as a first distribution update cycle, and every hour within every 8 hours as the update time corresponding to the second distribution update cycle within the first distribution cycle, that is, 0:00, 8:00, and 16:00 every day are the update times corresponding to the first distribution update cycle. During the three hours from 0:00 to 1:00, 8:00 to 9:00, and 16:00 to 17:00 every day, each electronic coupon to be issued is sorted only with offline virtual scores, so that the coupon can have a certain amount of time to cold start. At the same time, it avoids the situation where the user who receives the coupon in the first hour happens to use less, causing the coupon to be downgraded in subsequent sorting and unable to be continued to be issued, so as to further improve the accuracy of coupon issuance.

[0183] For example, if the conversion rate of a coupon is very poor, it will be downgraded between 1:00 and 8:00. However, since the first distribution update cycle will be updated at 8:00, that is, all coupons will be re-distributed for one hour at 8:00 with offline virtual scores, the downgraded coupon may be re-distributed. If the conversion rate of the coupon is relatively good between 8:00 and 9:00 and between 9:00 and 16:00 in the future, the coupon will not be downgraded, thus avoiding the situation where the coupon is downgraded on the day due to poor conversion rate in the first hour and is never recommended for distribution, thereby improving the rationality of coupon distribution.

[0184] In another exemplary implementation, for the newly submitted coupons to be issued on the day, the amount of receipt and the amount of use are both 0. At this time, the coupons have no real-time data, so an initial value of the target virtual score can be set, and the initial value can be the offline score calculated by the coupon offline score prediction model. In other words, for new coupons without real-time data, their offline scores can be used for sorting, and no downgrade is required. When there is real-time data, they can be sorted according to the target virtual score obtained by merging the offline virtual score and the real-time virtual score, and whether the coupon is a tail coupon can be determined based on the real-time data to determine whether it needs to be downgraded before participating in the sorting, and then the coupons are issued according to the sorting results.

[0185] In the present disclosure, through the first issuance cycle and the second issuance update cycle, the real-time data of the coupons can be fully considered, and the update times corresponding to the different issuance cycles of the coupons can be sorted and recommended in different ways according to the actual situation, thereby improving the accuracy and rationality of the coupon issuance.

[0186] Those skilled in the art will appreciate that all or part of the steps for implementing the above embodiments are implemented as a computer program executed by a CPU. When the computer program is executed by the CPU, the above functions defined by the above method provided by the present invention are performed. The program may be stored in a computer-readable storage medium, which may be a read-only memory, a disk, or an optical disk.

[0187] In addition, it should be noted that the above-mentioned figures are only schematic illustrations of the processes included in the method according to an exemplary embodiment of the present invention, and are not intended to be limiting. It is easy to understand that the processes shown in the above-mentioned figures do not indicate or limit the time sequence of these processes. In addition, it is also easy to understand that these processes can be performed synchronously or asynchronously, for example, in multiple modules.

[0188] Fig. 9 A schematic diagram showing the structure of an electronic coupon issuing device in an exemplary embodiment of the present disclosure is shown. Fig. 9The device may include an offline virtual score determination module 910, a real-time data acquisition module 920, a real-time virtual score determination module 930, a target virtual score determination module 940, and a coupon issuance module 950. Among them:

[0189] The offline virtual score determination module 910 is configured to input the attribute characteristics of the electronic coupon to be issued into the coupon offline score prediction model to obtain the offline virtual score of the electronic coupon to be issued;

[0190] The real-time data acquisition module 920 is configured to acquire the real-time collection amount and real-time usage amount of the electronic coupon to be issued from the last update time to the current time;

[0191] A real-time virtual score determination module 930 is configured to determine the real-time virtual score of the electronic coupon to be issued based on the real-time collection amount and the real-time usage amount and according to the Bayesian smoothing model corresponding to the electronic coupon to be issued;

[0192] The target virtual score determination module 940 is configured to determine the target virtual score of the electronic coupon to be issued according to the offline virtual score and the real-time virtual score;

[0193] The coupon issuing module 950 is configured to sort the electronic coupons to be issued based on the target virtual score, and issue the electronic coupons to be issued according to the sorting result.

[0194] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, the real-time data acquisition module 920 is specifically configured as follows:

[0195] Obtaining the real-time collection amount and real-time usage amount of the electronic coupon to be issued from the last update time corresponding to the first issuance update cycle to the current time;

[0196] Based on this, the coupon issuing module 950 is specifically configured as follows:

[0197] In a second issuance update cycle corresponding to the current moment, issuing the electronic coupons to be issued according to the sorting result;

[0198] The current time is the update time corresponding to the second issuance update period, and the first issuance update period includes at least one of the second issuance update periods.

[0199] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, the coupon offline score prediction model is predetermined in the following manner:

[0200] Acquire the attribute characteristics of the electronic coupons that have been issued and the conversion rate of the electronic coupons that have been issued;

[0201] Using the attribute features of the electronic coupons that have been issued as input samples and the conversion rate of the electronic coupons that have been issued as optimization target values, the gradient boosting iterative decision tree model is trained to obtain the offline prediction model of the coupons;

[0202] The conversion rate of the electronic coupon that has been issued is determined according to the ratio between the historical usage amount and the historical collection amount of the electronic coupon that has been issued.

[0203] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, the Bayesian smoothing model corresponding to the electronic coupon to be issued is determined by the following formula:

[0204]

[0205] Wherein, U is the real-time usage of the electronic coupon to be issued from the last update time corresponding to the first issuance update cycle to the current time, G is the real-time collection amount of the electronic coupon to be issued from the last update time corresponding to the first issuance update cycle to the current time, and β are smoothing parameters in the Bayesian smoothing model;

[0206] The smoothing parameter in the Bayesian smoothing model corresponding to each of the electronic coupons to be issued is determined in the following manner:

[0207] Acquire historical real-time data of the electronic coupon to be issued, the historical real-time data including the collection amount and usage amount of the electronic coupon to be issued between the update time of each first issuance cycle and the update time corresponding to each second issuance update cycle within the first issuance cycle within a recent preset time;

[0208] The historical real-time data is used as prior data, and based on moment estimation and maximum expectation algorithm, the smoothing parameters in the Bayesian smoothing model corresponding to the electronic coupon to be issued are obtained.

[0209] In some exemplary embodiments of the present disclosure, based on the aforementioned embodiments, the device 900 further includes a new coupon smoothing parameter determination module, which is specifically configured to: when the electronic coupon to be issued does not have the historical real-time data, determine that the electronic coupon to be issued is a new coupon; and determine the median or average of the smoothing parameters corresponding to other electronic coupons to be issued that have the historical real-time data as the smoothing parameters in the Bayesian smoothing model corresponding to the new coupon.

[0210] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, the step of sorting the to-be-issued electronic coupons based on the target virtual scores includes:

[0211] Determine the collection amount and usage amount of each electronic coupon to be issued from the last update time corresponding to the first issuance update cycle to the current time;

[0212] When the claimed amount is greater than a first preset threshold and the used amount is less than a second preset threshold, determining that the electronic coupon to be issued is a tail coupon; otherwise, determining that the electronic coupon to be issued is a normal coupon;

[0213] Based on the target virtual score, the normal coupons and the tail coupons are sorted in descending order to determine the sorting result of the electronic coupons to be issued;

[0214] Among them, in the sorting result, the sorting order of the normal ticket with the smallest target virtual score is placed before the sorting order of the tail ticket with the largest target virtual score.

[0215] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, the first preset threshold and the second preset threshold are determined in the following manner:

[0216] Obtain historical real-time data of each electronic coupon to be issued to generate a training sample set;

[0217] Performing unsupervised learning training on the clustering model according to the training sample set to perform binary classification on the electronic coupons to be issued;

[0218] Determining the first preset threshold and the second preset threshold according to the classification result;

[0219] The historical real-time data includes the collection amount and usage amount of the electronic coupons to be issued between the update time of each first issuance cycle and the update time corresponding to each second issuance update cycle within the first issuance cycle within the most recent preset time.

[0220] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, the apparatus 900 further includes an offline distribution module, which is specifically configured as follows:

[0221] In a target time period between an update time of each of the first issuance update cycles and an update time of a first second issuance update cycle within the first issuance update cycle, the electronic coupons to be issued are sorted in descending order based on the offline virtual scores, so that the electronic coupons to be issued are issued within the target time period according to the sorting result.

[0222] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, the target virtual score determination module 940 is specifically configured as follows:

[0223] Determining a first product between the offline virtual score and a first preset weight;

[0224] Determine a second product between the real-time virtual score and a second preset weight;

[0225] The target virtual score of the electronic coupon to be issued is determined according to the sum of the first product and the second product.

[0226] In some exemplary embodiments of the present disclosure, based on the foregoing embodiments, issuing the to-be-issued electronic coupons according to the ranking results includes:

[0227] According to the descending sorting result of the target virtual scores, the first K electronic coupons to be issued are selected from the electronic coupons to be issued for issuance, wherein K is a positive integer.

[0228] In some exemplary embodiments of the present disclosure, based on the aforementioned embodiments, the attribute characteristics include one or more of the category, denomination, limit, discount strength, popularity value of the covered inventory unit, price of the covered inventory unit, covered commodity category, covered commodity brand, and covered store of the electronic coupon.

[0229] The specific details of each unit in the above-mentioned electronic coupon issuing device have been described in detail in the corresponding electronic coupon issuing method, so they will not be repeated here.

[0230] It should be noted that, although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into multiple modules or units to be embodied.

[0231] In addition, although the steps of the method in the present disclosure are described in a specific order in the drawings, this does not require or imply that the steps must be performed in this specific order, or that all the steps shown must be performed to achieve the desired results. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step, and / or one step may be decomposed into multiple steps, etc.

[0232] Through the description of the above implementation, it is easy for those skilled in the art to understand that the example implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the implementation of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a mobile terminal, or a network device, etc.) to execute the method according to the implementation of the present disclosure.

[0233] In the exemplary embodiments of the present disclosure, a computer storage medium capable of implementing the above method is also provided. A program product capable of implementing the above method of this specification is stored thereon. In some possible embodiments, various aspects of the present disclosure may also be implemented in the form of a program product, which includes a program code, and when the program product is run on a terminal device, the program code is used to enable the terminal device to perform the steps according to various exemplary embodiments of the present disclosure described in the above "Exemplary Method" section of this specification.

[0234] refer to Fig.10 As shown, a program product 1000 for implementing the above method according to an embodiment of the present disclosure is described, which can adopt a portable compact disk read-only memory (CD-ROM) and include program code, and can be run on a terminal device, such as a personal computer. However, the program product of the present disclosure is not limited thereto. In this document, a readable storage medium can be any tangible medium containing or storing a program, which can be used by or in combination with an instruction execution system, an apparatus or a device.

[0235] The program product may use any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0236] Computer readable signal media may include data signals propagated in baseband or as part of a carrier wave, in which readable program code is carried. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Readable signal media may also be any readable medium other than a readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0237] The program code embodied on the readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination of the foregoing.

[0238] Program code for performing the operations of the present disclosure may be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving a remote computing device, the remote computing device may be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., through the Internet using an Internet service provider).

[0239] In addition, in an exemplary embodiment of the present disclosure, an electronic device capable of implementing the above method is also provided.

[0240] Those skilled in the art will appreciate that various aspects of the present disclosure may be implemented as systems, methods or program products. Therefore, various aspects of the present disclosure may be specifically implemented in the following forms, namely: complete hardware implementation, complete software implementation (including firmware, microcode, etc.), or a combination of hardware and software, which may be collectively referred to herein as "circuits", "modules" or "systems".

[0241] Refer to the following Fig.11 1100 according to this embodiment of the present disclosure is described. Fig.11 The electronic device 1100 shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0242] like Fig.11As shown, the electronic device 1100 is in the form of a general computing device. The components of the electronic device 1100 may include, but are not limited to: the at least one processing unit 1110, the at least one storage unit 1120, a bus 1130 connecting different system components (including the storage unit 1120 and the processing unit 1110), and a display unit 1140.

[0243] The storage unit stores a program code, which can be executed by the processing unit 1110, so that the processing unit 1110 performs the steps described in the "Exemplary Method" section of the present specification according to various exemplary embodiments of the present disclosure. For example, the processing unit 1110 can perform the following steps: Figure 1 As shown in: Step S110, input the attribute characteristics of the electronic coupon to be issued into the coupon offline score prediction model to obtain the offline virtual score of the electronic coupon to be issued; Step S120, obtain the real-time collection amount and real-time usage amount of the electronic coupon to be issued from the last update time to the current time; Step S130, based on the real-time collection amount and the real-time usage amount, according to the Bayesian smoothing model corresponding to the electronic coupon to be issued, determine the real-time virtual score of the electronic coupon to be issued. Step S140, determine the target virtual score of the electronic coupon to be issued according to the offline virtual score and the real-time virtual score; Step S150, sort the electronic coupons to be issued based on the target virtual score, and issue the electronic coupons to be issued according to the sorting result.

[0244] For another example, the processing unit 1110 may execute the following Figure 2 as well as Figures 4 to 8 Follow the steps shown in .

[0245] The storage unit 1120 may include a readable medium in the form of a volatile storage unit, such as a random access storage unit (RAM) 11201 and / or a cache storage unit 11202 , and may further include a read-only storage unit (ROM) 11203 .

[0246] The storage unit 1120 may also include a program / utility 11204 having a set (at least one) of program modules 11205, such program modules 11205 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0247] Bus 1130 may represent one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

[0248] The electronic device 1100 may also communicate with one or more external devices 1200 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), may also communicate with one or more devices that enable a user to interact with the electronic device 1100, and / or communicate with any device that enables the electronic device 1100 to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed via an input / output (I / O) interface 1150. Furthermore, the electronic device 1100 may also communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via a network adapter 1160. As shown, the network adapter 1160 communicates with other modules of the electronic device 1100 via a bus 1130. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 1100, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0249] Through the description of the above implementation, it is easy for those skilled in the art to understand that the example implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the implementation of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method according to the implementation of the present disclosure.

[0250] In addition, the above-mentioned figures are only schematic illustrations of the processes included in the method according to the exemplary embodiments of the present disclosure, and are not intended to be limiting. It is easy to understand that the processes shown in the above-mentioned figures do not indicate or limit the time sequence of these processes. In addition, it is also easy to understand that these processes can be performed synchronously or asynchronously, for example, in multiple modules.

[0251] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary technical means in the art that are not disclosed in the present disclosure. The specification and examples are to be considered as exemplary only, and the true scope and spirit of the present disclosure are indicated by the claims.

Claims

1. A method for issuing electronic coupons, It is characterized in that include: Inputting the attribute characteristics of the electronic coupon to be issued into the coupon offline score prediction model to obtain the offline virtual score of the electronic coupon to be issued; Obtaining the real-time collection amount and real-time usage amount of the electronic coupon to be issued from the last update time to the current time; Based on the real-time collection amount and the real-time usage amount, and according to the Bayesian smoothing model corresponding to the electronic coupon to be issued, determining the real-time virtual score of the electronic coupon to be issued; Determining a target virtual score of the electronic coupon to be issued according to the offline virtual score and the real-time virtual score; sorting the electronic coupons to be issued based on the target virtual scores, and issuing the electronic coupons to be issued according to the sorting results; The Bayesian smoothing model corresponding to the electronic coupon to be issued is determined by the following formula: Wherein, U is the real-time usage of the electronic coupon to be issued from the last update time corresponding to the first issuance update cycle to the current time, G is the real-time collection amount of the electronic coupon to be issued from the last update time corresponding to the first issuance update cycle to the current time, and β are smoothing parameters in the Bayesian smoothing model; Wherein, the smoothing parameters in the Bayesian smoothing model corresponding to each of the electronic coupons to be issued are determined in the following manner: acquiring historical real-time data of the electronic coupons to be issued, the historical real-time data including the collection amount and usage amount of the electronic coupons to be issued between the update time of each first issuance cycle and the update time corresponding to each second issuance update cycle within the first issuance cycle within the most recent preset time; using the historical real-time data as prior data, based on moment estimation and maximum expectation algorithm, obtaining the smoothing parameters in the Bayesian smoothing model corresponding to the electronic coupons to be issued; the first issuance update cycle includes at least one of the second issuance update cycles.

2. The electronic coupon distribution method according to claim 1, It is characterized in that The obtaining of the real-time collection amount and real-time usage amount of the electronic coupon to be issued from the last update time to the current time includes: Obtaining the real-time collection amount and real-time usage amount of the electronic coupon to be issued from the last update time corresponding to the first issuance update cycle to the current time; The issuing of the electronic coupons to be issued according to the sorting results includes: In a second issuance update cycle corresponding to the current moment, issuing the electronic coupons to be issued according to the sorting result; The current time is the update time corresponding to the second issuance update cycle.

3. The electronic coupon distribution method according to claim 1, It is characterized in that The coupon offline score prediction model is predetermined by: Acquire the attribute characteristics of the electronic coupons that have been issued and the conversion rate of the electronic coupons that have been issued; Using the attribute features of the electronic coupons that have been issued as input samples and the conversion rate of the electronic coupons that have been issued as optimization target values, the gradient boosting iterative decision tree model is trained to obtain the offline prediction model of the coupons; The conversion rate of the electronic coupon that has been issued is determined according to the ratio between the historical usage amount and the historical collection amount of the electronic coupon that has been issued.

4. The electronic coupon distribution method according to claim 1, It is characterized in that The method further comprises: When the electronic coupon to be issued does not have the historical real-time data, determining that the electronic coupon to be issued is a new coupon; The median or average of the smoothing parameters corresponding to other electronic coupons to be issued that have the historical real-time data is determined as the smoothing parameter in the Bayesian smoothing model corresponding to the new coupon.

5. The electronic coupon distribution method according to claim 2, It is characterized in that The step of sorting the electronic coupons to be issued based on the target virtual scores includes: Determine the collection amount and usage amount of each electronic coupon to be issued from the last update time corresponding to the first issuance update cycle to the current time; When the claimed amount is greater than a first preset threshold and the used amount is less than a second preset threshold, determining that the electronic coupon to be issued is a tail coupon; otherwise, determining that the electronic coupon to be issued is a normal coupon; Based on the target virtual score, the normal coupons and the tail coupons are sorted in descending order to determine the sorting result of the electronic coupons to be issued; Among them, in the sorting result, the sorting order of the normal ticket with the smallest target virtual score is placed before the sorting order of the tail ticket with the largest target virtual score.

6. The electronic coupon distribution method according to claim 5, It is characterized in that The first preset threshold and the second preset threshold are determined in the following manner: Obtain historical real-time data of each electronic coupon to be issued to generate a training sample set; Performing unsupervised learning training on the clustering model according to the training sample set to perform binary classification on the electronic coupons to be issued; Determining the first preset threshold and the second preset threshold according to the classification result; The historical real-time data includes the collection amount and usage amount of the electronic coupons to be issued between the update time of each first issuance cycle and the update time corresponding to each second issuance update cycle within the first issuance cycle within the most recent preset time.

7. The electronic coupon distribution method according to claim 2, It is characterized in that The method further comprises: In a target time period between an update time of each of the first issuance update cycles and an update time of a first second issuance update cycle within the first issuance update cycle, the electronic coupons to be issued are sorted in descending order based on the offline virtual scores, so that the electronic coupons to be issued are issued within the target time period according to the sorting result.

8. The electronic coupon distribution method according to claim 1, It is characterized in that The step of determining the target virtual score of the electronic coupon to be issued according to the offline virtual score and the real-time virtual score comprises: Determining a first product between the offline virtual score and a first preset weight; Determine a second product between the real-time virtual score and a second preset weight; The target virtual score of the electronic coupon to be issued is determined according to the sum of the first product and the second product.

9. The electronic coupon distribution method according to claim 1, It is characterized in that The issuing of the electronic coupons to be issued according to the ranking results includes: According to the descending sorting result of the target virtual scores, the first K electronic coupons to be issued are selected from the electronic coupons to be issued for issuance, wherein K is a positive integer.

10. The electronic coupon distribution method according to any one of claims 1 to 9, It is characterized in that The attribute features include one or more of the category, denomination, limit, discount strength, popularity value of the covered inventory unit, price of the covered inventory unit, covered commodity category, covered commodity brand, and covered store of the electronic coupon.

11. An electronic coupon issuing device, It is characterized in that include: An offline virtual score determination module is configured to input the attribute characteristics of the electronic coupon to be issued into the coupon offline score prediction model to obtain the offline virtual score of the electronic coupon to be issued; A real-time data acquisition module is configured to acquire the real-time collection amount and real-time usage amount of the electronic coupon to be issued from the last update time to the current time; A real-time virtual score determination module is configured to determine the real-time virtual score of the electronic coupon to be issued based on the real-time collection amount and the real-time usage amount and according to the Bayesian smoothing model corresponding to the electronic coupon to be issued; a target virtual score determination module, configured to determine the target virtual score of the electronic coupon to be issued according to the offline virtual score and the real-time virtual score; A coupon issuing module, configured to sort the electronic coupons to be issued based on the target virtual score, and issue the electronic coupons to be issued according to the sorting result; The Bayesian smoothing model corresponding to the electronic coupon to be issued is determined by the following formula: Wherein, U is the real-time usage of the electronic coupon to be issued from the last update time corresponding to the first issuance update cycle to the current time, G is the real-time collection amount of the electronic coupon to be issued from the last update time corresponding to the first issuance update cycle to the current time, and β are smoothing parameters in the Bayesian smoothing model; Wherein, the smoothing parameters in the Bayesian smoothing model corresponding to each of the electronic coupons to be issued are determined in the following manner: acquiring historical real-time data of the electronic coupons to be issued, the historical real-time data including the collection amount and usage amount of the electronic coupons to be issued between the update time of each first issuance cycle and the update time corresponding to each second issuance update cycle within the first issuance cycle within the most recent preset time; using the historical real-time data as prior data, based on moment estimation and maximum expectation algorithm, obtaining the smoothing parameters in the Bayesian smoothing model corresponding to the electronic coupons to be issued; the first issuance update cycle includes at least one of the second issuance update cycles.

12. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the program is executed by a processor, the electronic coupon issuing method according to any one of claims 1 to 10 is implemented.

13. An electronic device, It is characterized in that include: one or more processors; A storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the electronic coupon issuing method according to any one of claims 1 to 10.

Images