Method, device and equipment for pushing rights and interests and medium

Abstract

The embodiment of the invention discloses a method and device for pushing rights and interests, equipment and a medium. According to the scheme, the method comprises the following steps: after order information of a to-be-processed order of a user is obtained, determining various staging forms of staging services corresponding to to-be-processed amount information included in the order information, and for a plurality of candidate rights and interests sequences, utilizing a conversion rate prediction model to predict a conversion rate of the to-be-processed amount information; and predicting conversion rates of the user for various staging forms under the condition of each candidate right sequence, determining a target right sequence meeting a preset conversion rate condition from the plurality of candidate right sequences based on the conversion rates, determining target right information corresponding to various staging forms according to the target right sequence, and sending the target right information to the user. And pushing the target right and interest information to the user.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method, device, equipment, and medium for pushing rights and interests. Background Art

[0002] With the development of the times, installment payments have become an important payment method for consumers to purchase goods and services. To promote the development of installment payment services, major financial institutions and e-commerce platforms have launched various installment payment promotions. For example, installment payment services can promote benefits to users, such as coupons, credit points, and other benefits.

[0003] In view of this, it is necessary to provide a high-precision rights push solution to reduce the waste of network resources or computer resources. Summary of the Invention

[0004] In view of this, embodiments of the present application provide a method, apparatus, device, and medium for pushing rights, so as to provide a high-precision rights pushing solution.

[0005] To solve the above technical problems, the embodiments of this specification provide a method for pushing rights and interests, including:

[0006] Obtaining order information of a user's pending order; the order information includes information on the amount to be processed;

[0007] Determine various installment forms of the installment business corresponding to the amount information;

[0008] For several candidate benefit sequences, a conversion rate prediction model is used to predict the user's conversion rate for the various installment forms under the conditions of each candidate benefit sequence. The number of candidate benefits included in a candidate benefit sequence is the same as the number of installment forms, and each installment form corresponds to one candidate benefit. For any two installment forms among the various installment forms, the candidate benefit corresponding to the installment form with a longer total installment duration has a higher equity value.

[0009] Based on the conversion rate, determining a target equity sequence that meets a preset conversion rate condition from the plurality of candidate equity sequences;

[0010] According to the target equity sequence, determining target equity information corresponding to the various installment forms;

[0011] Push the target rights and interests information to the user.

[0012] The embodiments of this specification also provide a device for pushing rights and interests, including:

[0013] An information acquisition module is used to obtain order information of a user's pending order; the order information includes information on the amount to be processed;

[0014] A first determining module is used to determine various installment forms of the installment business corresponding to the amount information;

[0015] A model prediction module is configured to use a conversion rate prediction model to predict, for a plurality of candidate benefit sequences, the user's conversion rate for the various installment forms under the conditions of each candidate benefit sequence; the number of candidate benefits included in a candidate benefit sequence is equal to the number of types of installment forms, and each installment form corresponds to one candidate benefit; for any two installment forms among the various installment forms, the candidate benefit corresponding to the installment form with a longer total installment duration has a higher equity value;

[0016] A second determining module is configured to determine, based on the conversion rate, a target equity sequence that meets a preset conversion rate condition from the plurality of candidate equity sequences;

[0017] A third determining module is configured to determine target equity information corresponding to the various installment forms according to the target equity sequence;

[0018] The information push module is used to push the target rights and interests information to the user.

[0019] The embodiments of this specification also provide a device for pushing benefits, including:

[0020] at least one processor; and,

[0021] a memory communicatively connected to the at least one processor; wherein,

[0022] The memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to:

[0023] Obtaining order information of a user's pending order; the order information includes information on the amount to be processed;

[0024] Determine various installment forms of the installment business corresponding to the amount information;

[0025] For several candidate benefit sequences, a conversion rate prediction model is used to predict the user's conversion rate for the various installment forms under the conditions of each candidate benefit sequence. The number of candidate benefits included in a candidate benefit sequence is the same as the number of installment forms, and each installment form corresponds to one candidate benefit. For any two installment forms among the various installment forms, the candidate benefit corresponding to the installment form with a longer total installment duration has a higher equity value.

[0026] Based on the conversion rate, determining a target equity sequence that meets a preset conversion rate condition from the plurality of candidate equity sequences;

[0027] According to the target equity sequence, determining target equity information corresponding to the various installment forms;

[0028] Push the target rights and interests information to the user.

[0029] The embodiments of this specification also provide a computer-readable storage medium storing computer-executable instructions, which implement the above-mentioned method of pushing benefits when executed by a processor.

[0030] At least one embodiment provided in this specification can achieve the following beneficial effects:

[0031] In the embodiments of this specification, after obtaining the order information of a user's pending order, the various installment types of the installment business corresponding to the pending amount information included in the order information can be determined. Using a conversion rate prediction model, the user's conversion rate for each installment type under the conditions of each candidate benefit sequence can be predicted. Furthermore, based on the conversion rate, target benefit information corresponding to a target benefit sequence that meets the preset conversion rate conditions is determined from a number of candidate benefit sequences and pushed to the user. Thus, using a model to predict conversion rates based on the user's pending order information can be more rapid and accurate, requiring less human intervention, and thus improving the efficiency of target benefit determination without excessive use of human resources.

[0032] On the other hand, by limiting the candidate rights and interests corresponding to installment forms with longer total installment durations to higher rights and interests values, accurate distribution of rights and interests can be achieved, which is conducive to increasing users' willingness to use rights and interests, improving conversion rates, and thus saving network or server computing resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the embodiments of this specification or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments recorded in this application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0034] Figure 1 A schematic diagram of an application scenario of a method for pushing benefits provided in an embodiment of this specification;

[0035] Figure 2 A flowchart of a method for pushing benefits provided in an embodiment of this specification;

[0036] Figure 3 A schematic diagram of a terminal interface displaying target equity information provided in an embodiment of this specification;

[0037] Figure 4 A flowchart of a conversion rate prediction model training method provided in an embodiment of this specification;

[0038] Figure 5 This is an overall flow chart of a method for pushing benefits provided in an embodiment of this specification;

[0039] Figure 6 The embodiments of this specification provide corresponding Figure 2 A schematic diagram of the structure of a device for pushing rights;

[0040] Figure 7 This is a schematic diagram of the structure of a device for pushing benefits provided in an embodiment of this specification. DETAILED DESCRIPTION

[0041] The following description sets forth many specific details to facilitate a thorough understanding of the present application. However, the present application can be implemented in many other ways than those described herein, and those skilled in the art can make similar generalizations without violating the scope of the present application. Therefore, the present application is not limited to the specific implementations disclosed below.

[0042] The terms used in one or more embodiments of the present application are for the purpose of describing specific embodiments only and are not intended to limit one or more embodiments of the present application. The singular forms "a", "the" and "the" used in one or more embodiments of the present application and the appended claims are also intended to include plural forms, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in one or more embodiments of the present application refers to and includes any or all possible combinations of one or more associated listed items.

[0043] It should be understood that although the terms first, second, etc. may be used to describe various information in one or more embodiments of the present application, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from each other. For example, without departing from the scope of one or more embodiments of the present application, the first may also be referred to as the second, and similarly, the second may also be referred to as the first. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0044] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of the relevant regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0045] In the prior art, installment payment services usually have installment repayment periods of 3, 6, 12, or 24 installments. For example, 3 installments can mean that the user needs to repay the total amount in 3 installments, and the total installment duration can be 3 months, with partial repayment each month, and the entire payment is paid off in 3 months. Providing users with appropriate discounts is conducive to users choosing suitable installment services. In actual applications, servers usually randomly push discount information to users, and the discount information may not meet the user's needs. This may result in the user being pushed discount information, but the user not using it, resulting in a waste of network resources or server resources, and may also cause some inconvenience to the user.

[0046] In order to solve the defects in the prior art, this solution provides the following embodiments.

[0047] Figure 1 A schematic diagram of an application scenario of a method for pushing benefits provided in an embodiment of this specification.

[0048] like Figure 1 As shown, after server 101 obtains the order information of a user's pending order, it can determine the various installment types of the installment business corresponding to the pending amount information included in the order information, and determine several candidate benefit sequences based on these various installment types. Subsequently, a conversion rate prediction model can be used to predict the user's conversion rate for each candidate benefit sequence under the conditions of each candidate benefit sequence. Based on these conversion rates, a target benefit sequence that meets the preset conversion rate conditions is determined from the candidate benefit sequences. According to these target benefit sequences, target benefit information corresponding to the various installment types is determined, and finally, this target benefit information is pushed to the user's terminal device 102.

[0049] In actual applications, server 101 can have a rights push function. After server 101 obtains user feature data and order feature data, it can use the conversion rate prediction model to process these data in real time, determine the target rights information, and push the target rights information to the user to improve the user conversion rate.

[0050] The user's terminal device 102 includes but is not limited to a smart phone, tablet computer, PDA, smart watch, etc. The server 101 includes but is not limited to any device, equipment, platform, server cluster, etc. with computing and processing capabilities.

[0051] Figure 2 This is a flow chart of a method for pushing benefits provided in an embodiment of this specification. From a hardware perspective, the execution subject of this process can be a server. From a program perspective, the execution subject of this process can be a program installed on the server. Figure 2 As shown, the process may include the following steps:

[0052] Step 202: Obtain order information of the user's pending order; the order information includes the amount information to be processed.

[0053] In the embodiments of this specification, the order information of the user's pending order may include but is not limited to: pending order amount information, order identification information, order placement time information, order payee information, order payee information, order completion time limit information, etc.

[0054] In practical applications, a user's pending order can be a pending payment order for a product or service, or it can be an order for a user to repay a loan after completing a loan on a lending platform, without specific limitations. For pending orders, users can choose to pay in installments. If a user wishes to pay in installments, they can select one or more installment payment platforms to conduct the installment payment service.

[0055] In practice, installment payment services offered by various platforms typically include multiple installment repayment periods. For example, Platform A offers installment payment products including 3-installment products, 6-installment products, 12-installment products, and 24-installment products. Another example: Platform B offers installment payment products including 6-installment products, 12-installment products, 24-installment products, and 36-installment products. Different platforms can set different installment periods and corresponding repayment interest rates based on business needs. Users can choose the installment payment platform and installment repayment period based on their needs.

[0056] Step 204: Determine various installment forms of the installment business corresponding to the amount information.

[0057] In the embodiments of this specification, the installment form of the installment business can represent the form information of the number of installments supported by the installment business. The installment form of the installment business can be set on a monthly basis, such as: 3 installments, 6 installments, 9 installments, 12 installments, 24 installments, etc., wherein the time interval between each installment is one month. In actual applications, the time interval between each installment can also be one quarter or one year, and there is no specific limitation on the number of installments and the time interval between each installment. The business issuer of the installment business can set and adjust the installment form according to demand, and there is no specific limitation on the installment form of the installment business.

[0058] In actual applications, installment service providers can also offer different installment plans based on the order amount. For example, for orders less than 10,000 yuan, installment plans such as 3, 6, 9, or 12 installments can be pushed; for orders greater than 10,000 yuan, installment plans such as 12, 24, 36, or 48 installments can be pushed. Thus, after obtaining the order information for a user's pending order, the installment plans corresponding to that amount can be determined based on the pending amount included in the order information. For example, if the pending amount included in the order information for user A's pending order is 6,000 yuan, which is less than 10,000 yuan, the corresponding installment plans can be determined as 3, 6, 9, or 12 installments. Another example: if the pending amount included in the order information for user B's pending order is 30,000 yuan, which is greater than 10,000 yuan, the corresponding installment plans can be determined as 12, 24, 36, or 48 installments. It should be noted that the numerical values ​​in the above examples are examples and should not limit this application. In addition, in actual applications, for different order amounts, the business issuer of the installment business can also provide the same installment form.

[0059] Step 206: For several candidate benefit sequences, using a conversion rate prediction model, predict the user's conversion rate for the various installment forms under the conditions of each candidate benefit sequence.

[0060] Among them, the number of candidate interests included in a candidate interest sequence is the same as the number of types of the installment forms, and one installment form corresponds to one candidate interest; for any two installment forms among the various installment forms, the candidate interest corresponding to the installment form with a longer total installment duration has a higher equity value.

[0061] In the embodiments of the present specification, the conversion rate prediction model can be a model obtained after training of a neural network model, wherein the neural network model is a computational model that imitates the structure and function of a biological neural network and is widely used in the fields of artificial intelligence, machine learning, and deep learning. The neural network model consists of a large number of artificial neurons (nodes), which are connected by weights and can learn and process complex data. The structure of the neural network model can generally be divided into an input layer, a hidden layer, and an output layer, wherein the input layer is used to receive the raw input data, the hidden layer is used to extract the features of the data, and the output layer is used to generate the final prediction result or classification label.

[0062] In the embodiments of this specification, the conversion rate prediction model may include but is not limited to: Convolutional Neural Networks (CNN) model, Recurrent Neural Network (RNN) model, Bidirectional Recurrent Neural Network (Bi-RNN) model, Long Short-Term Memory (LSTM) model, Gated Recurrent Unit (GRU) model, etc.

[0063] In actual application, the conversion rate prediction model can be obtained by training a neural network classification model using training data, wherein the training data may include historical business data of historical users who have obtained historical rights and interests information and user feature data of users; the historical rights and interests information is rights and interests information determined based on the several candidate rights and interests sequences; the historical business data includes historical order information of the historical users, the historical rights and interests information, and information on whether the historical users have handled installment business based on the historical rights and interests information. The neural network classification model can learn the characteristics of the data through the training process. After training, the neural network classification model can predict the conversion rate of users for each installment form under the conditions of the candidate rights and interests sequence based on the candidate rights and interests sequence, user feature data and order data. The training process of the model can generally include steps such as forward propagation, loss calculation, backpropagation and weight update. Since the subsequent embodiments in the embodiments of this specification will explain the training process of the model in detail, it will not be repeated here.

[0064] In the embodiments of this specification, candidate benefits may include, but are not limited to, credit points, discount coupons, discount coupons, deduction coupons, interest-free coupons, instant discount red envelopes, and cashback red envelopes. For example, if the candidate benefit is credit points, the corresponding equity value of the candidate benefit may be the credit point value. For another example, if the candidate benefit is a voucher, the corresponding equity value of the candidate benefit may be the voucher deduction amount. For another example, if the candidate benefit is a red envelope, the corresponding equity value of the candidate benefit may be the red envelope amount.

[0065] In the embodiments of this specification, one installment form corresponds to one candidate interest, and the number of candidate interests included in a candidate interest sequence is the same as the number of types of the installment form. For example: if the number of types of installment forms is 4, namely: 3 periods, 6 periods, 9 periods and 12 periods, then the number of candidate interests included in a candidate interest sequence is also 4, and 3 periods corresponds to one candidate interest, 6 periods corresponds to one candidate interest, 9 periods corresponds to one candidate interest, and 12 periods corresponds to one candidate interest. For example: if the candidate interest is a bond, a candidate interest sequence is [1, 2, 3, 4], which can indicate that the preferential amount corresponding to the 3-period installment is 1 yuan, the preferential amount corresponding to the 6-period installment is 2 yuan, the preferential amount corresponding to the 9-period installment is 3 yuan, and the preferential amount corresponding to the 12-period installment is 4 yuan. For example, if the candidate equity is credit points, a candidate equity sequence of [10, 20, 30, 40] can indicate that the credit point value corresponding to a 3-period installment is 10, the credit point value corresponding to a 6-period installment is 20, the credit point value corresponding to a 9-period installment is 30, and the credit point value corresponding to a 12-period installment is 40.

[0066] In the embodiments of this specification, for any two installment forms among various installment forms, the installment form with a longer total installment duration has a higher equity value for the candidate equity. The following is an exemplary explanation of the case where the candidate equity is credit points. For example, there are four types of installment forms, namely: 3 installments, 6 installments, 9 installments, and 12 installments. From 3 installments to 12 installments, the total installment duration increases successively. The credit point value corresponding to 6 installments can be greater than the credit point value corresponding to 3 installments, the credit point value corresponding to 9 installments can be greater than the credit point value corresponding to 6 installments, and the credit point value corresponding to 12 installments can be greater than the credit point value corresponding to 9 installments. This can encourage users to use the pushed equity.

[0067] In actual applications, multiple types of candidate benefits can be set, and different installment types can correspond to candidate benefits of different amounts. Of course, at least some of the different installment types can also correspond to candidate benefits of the same amount. The following example uses the case where the candidate benefits are credit points. For example, there are four types of installment types: 3 installments, 6 installments, 9 installments, and 12 installments. The preset credit point values ​​are: 10 credit points, 20 credit points, 30 credit points, 40 credit points, 50 credit points, and 60 credit points. The candidate benefit sequence can include, but is not limited to: [10, 20, 30, 40], [20, 30, 40, 50], [30, 40, 50, 60], etc., where the first digit in the candidate benefit sequence represents the credit point value corresponding to 3 installments, the second digit represents the credit point value corresponding to 6 installments, the third digit represents the credit point value corresponding to 9 installments, and the fourth digit represents the credit point value corresponding to 12 installments.

[0068] Step 208: Based on the conversion rate, determine a target equity sequence that meets a preset conversion rate condition from the plurality of candidate equity sequences.

[0069] In the embodiments of this specification, the conversion rate may refer to the probability that a user will select a certain installment form under the conditions of a certain candidate equity sequence. Using the conversion rate prediction model, the user's conversion rate for various installment forms under the conditions of a certain candidate equity sequence can be predicted. For example, there are four types of installment forms, namely: 3 periods, 6 periods, 9 periods, and 12 periods. Assuming that the candidate equity sequence is [20, 30, 40, 50], the conversion rate prediction model predicts that under the conditions of the candidate equity sequence, the user's conversion rate for 3 periods is predicted to be 0.2, the conversion rate for 6 periods is 0.3, the conversion rate for 9 periods is 0.2, and the conversion rate for 12 periods is 0.1. Among them, the candidate equity sequence [2, 3, 4, 5] can indicate that the user can obtain 20 credit points by choosing 3 periods, 30 credit points by choosing 6 periods, 40 credit points by choosing 9 periods, and 50 credit points by choosing 12 periods. A conversion rate of 0.2 for 3 periods indicates that the probability of the user choosing 3 periods is 0.2, a conversion rate of 0.3 for 6 periods indicates that the probability of the user choosing 6 periods is 0.3, a conversion rate of 0.2 for 9 periods indicates that the probability of the user choosing 9 periods is 0.2, and a conversion rate of 0.1 for 12 periods indicates that the probability of the user choosing 12 periods is 0.1.

[0070] In practical applications, the conversion rate prediction model can also be used to predict the user's conversion rate under the conditions of a certain candidate equity sequence. The user's conversion rate under the conditions of a certain candidate equity sequence can refer to the probability that the user chooses installment payment under the conditions of a certain candidate equity sequence.

[0071] In an embodiment of the present specification, the preset conversion rate condition may be a condition pre-set by the equity issuer or the equity service provider, for example: the preset conversion rate condition may be a candidate equity sequence with the highest user conversion rate among several candidate equity sequences, and accordingly, the candidate equity sequence with the highest user conversion rate may be determined as the target equity sequence; or, the preset conversion rate condition may also be a candidate equity sequence with a user conversion rate reaching a preset threshold, and accordingly, the candidate equity sequence with a user conversion rate reaching the preset threshold may be determined as the target equity sequence; or, the preset conversion rate condition may also be a candidate equity sequence with the highest sum of conversion rates of various installment forms, and accordingly, the candidate equity sequence with the highest sum of conversion rates of various installment forms may be determined as the target equity sequence; the preset conversion rate condition may be set and adjusted according to business needs, and there is no specific limitation on the preset conversion rate condition.

[0072] Step 210: Determine the target equity information corresponding to the various installment forms according to the target equity sequence.

[0073] In the embodiments of this specification, since a single equity sequence can determine the equity information corresponding to each installment format, once the target equity sequence is determined, the target equity information corresponding to each installment format can be determined. For example, if there are four installment formats: 3 installments, 6 installments, 9 installments, and 12 installments, and the target equity sequence is [20, 30, 40, 50], then the target equity corresponding to 3 installments can be determined to be 20 credit points, the target equity corresponding to 6 installments to be 30 credit points, the target equity corresponding to 9 installments to be 40 credit points, and the target equity corresponding to 12 installments to be 50 credit points.

[0074] In the embodiments of this specification, the target benefit information may include but is not limited to: the amount of the target benefit, the eligibility for receiving the target benefit, the time limit for receiving the target benefit, the type of the target benefit, etc.

[0075] Step 212: Push the target rights and interests information to the user.

[0076] In an embodiment of the present specification, after the server determines the target benefit information corresponding to various installment types, it can send the target benefit information to the user's terminal device. After the user's terminal device receives the target benefit information, it can display the target benefit information corresponding to the various installment types, so that the user can select a specific installment type based on the target benefit information corresponding to the various installment types. For example, when the benefit is credit points, the installment types are: 3 installments, 6 installments, 9 installments, and 12 installments, and the target benefit sequence is [20, 30, 40, 50]. After the user's terminal device receives the corresponding target benefit information, it can display on the screen: 20 credit points can be obtained in 3 installments, 30 credit points can be obtained in 6 installments, 40 credit points can be obtained in 9 installments, and 50 credit points can be obtained in 12 installments.

[0077] For example, Figure 3 This is a schematic diagram of a terminal interface displaying target rights information provided by an embodiment of this specification, such as Figure 3As shown, the terminal interface can display the order amount, such as "￥666.66." This page can also display installment information provided by various installment service providers. For example, Bank M's credit card installment service is divided into "3 installments," "6 installments," "9 installments," and "12 installments." The interface also displays the required repayment amount for each installment, the interest amount for each installment, and the benefits available. For example, "3 installments" offers "20 credit points," "6 installments" offers "30 credit points," "9 installments" offers "40 credit points," and "12 installments" offers "50 credit points." Bank N's credit card installment service is divided into "3 installments," "6 installments," "12 installments," and "24 installments." For example, "3 installments" offers "interest-free," "6 installments" offers "30 credit points," "12 installments" offers "60 credit points," and "24 installments" offers "90 credit points." The user can choose to apply for an installment plan with Bank M or Bank N. For example, the user can click to select the desired installment plan. This interface may also include a confirmation control, such as a "Confirm Payment" control. After selecting the desired installment plan, the user can click "Confirm Payment" to complete the order and installment plan. This page may also display the user's account information, such as "49*******." To facilitate the user's understanding of the installment plan, the installment plan interface may also display information such as the estimated repayment amount for each installment based on the order amount, various installment types, and interest rates. It may also display information such as interest payments. In actual applications, users may also choose to pay without installments. The interface may also include a control for selecting non-installment payment. The above interface is merely an example. In actual applications, not all of the content in the example may be displayed. The specific interface content can be customized based on actual needs and is not specifically defined here.

[0078] Figure 2 After obtaining the order information of a user's pending order, the method can determine the various installment types of the installment business corresponding to the pending amount information included in the order information. Using a conversion rate prediction model, the user's conversion rate for each installment type under the conditions of various candidate benefit sequences is predicted. Furthermore, based on the conversion rate, target benefit information corresponding to a target benefit sequence that meets the preset conversion rate conditions is determined from a number of candidate benefit sequences and pushed to the user. This allows the model to predict conversion rates based on the user's pending order information, resulting in faster and more accurate results without excessive human intervention, which helps improve the efficiency of target benefit determination and reduces the use of excessive human resources.

[0079] On the other hand, by limiting the candidate rights and interests corresponding to installment forms with longer total installment durations to higher rights and interests values, accurate distribution of rights and interests can be achieved, which is conducive to increasing users' willingness to use rights and interests, improving conversion rates, and thus saving network or server computing resources.

[0080] based on Figure 2 The method in this specification also provides some specific implementation plans of the method, which are described below.

[0081] Optional, Figure 2 The method may further include:

[0082] Sort each candidate interest by its value to obtain sorted candidate interests;

[0083] A preset number of candidate equities are selected from the sorted candidate equities in the sorting order to obtain a candidate equities sequence; the preset number is equal to the number represented by the number of types in the installment form.

[0084] In this embodiment of the present specification, the number of candidate rights included in a candidate rights sequence is the same as the number of installment types. For example, if there are three installment types, the number of candidate rights included in a candidate rights sequence is three; if there are four installment types, the number of candidate rights included in a candidate rights sequence is four.

[0085] In actual applications, the total installment duration corresponding to the first candidate interest in a candidate interest sequence to the total installment duration corresponding to the last candidate interest can be increasing or decreasing in sequence. In order to increase the user's willingness to choose the interest, the equity value of the candidate interest corresponding to the installment form with a longer total installment duration can be higher. In the embodiment of this specification, if the total installment duration corresponding to the first candidate interest in the candidate interest sequence to the total installment duration corresponding to the last candidate interest is increasing, each candidate interest can be sorted in order of equity value from low to high, and a preset number of candidate interests can be selected from the sorted candidate interests in the sorted order. If the total installment duration corresponding to the first candidate interest in the candidate interest sequence to the total installment duration corresponding to the last candidate interest is decreasing, each candidate interest can be sorted in order of equity value from high to low, and a preset number of candidate interests can be selected from the sorted candidate interests in the sorted order.

[0086] In practical applications, the determination of candidate benefit sequences can be independent of the pending amount information in the order information. Thus, the step of selecting a preset number of candidate benefits from the sorted candidate benefits in order of sorting and generating a candidate benefit sequence can be performed before obtaining the user's pending order information. The step of generating a candidate benefit sequence can be performed offline. For different orders with the same number of installment types, the server can use the same pre-determined candidate benefit sequences and select the appropriate candidate benefit sequence from each candidate benefit sequence for each order.

[0087] For example, if there are four types of installment plans, namely 3 installments, 6 installments, 9 installments, and 12 installments, then the corresponding preset number is 4. Assuming the candidate benefits are credit points, and the credit point values ​​are 10 credit points, 20 credit points, 30 credit points, 40 credit points, 50 credit points, and 60 credit points, these credit points are sorted from smallest to largest by credit point value, and four candidate benefits are selected from them in the sorted order to obtain a candidate benefit sequence. The selected candidate benefit sequences may include, but are not limited to, [10, 20, 30, 40], [20, 30, 40, 50], [30, 40, 50, 60], [20, 30, 50, 60], [10, 30, 40, 60], etc., where the first digit in the candidate benefit sequence represents the credit point value corresponding to 3 installments, the second digit represents the credit point value corresponding to 6 installments, the third digit represents the credit point value corresponding to 9 installments, and the fourth digit represents the credit point value corresponding to 12 installments.

[0088] Optionally, selecting a preset number of candidate interests from the sorted candidate interests in order of sorting to obtain a candidate interest sequence may specifically include:

[0089] The preset number of consecutive candidate rights are selected from the sorted candidate rights in the sorted order to obtain a candidate rights sequence.

[0090] In actual applications, a preset number of candidate interests are selected from the sorted candidate interests in the sorted order. The selection can be continuous or discontinuous. For example, assuming that the candidate interests are credit points, and the credit point values ​​are divided into: 10 credit points, 20 credit points, 30 credit points, 40 credit points, 50 credit points, and 60 credit points, and a candidate interest sequence contains 4 candidate interests, then the candidate interest sequences [10, 20, 30, 40], [20, 30, 40, 50], and [30, 40, 50, 60] are candidate interest sequences obtained by selecting consecutive candidate interests in the sorted order. Candidate interest sequences such as [20, 30, 50, 60], [10, 30, 40, 60], and [10, 20, 50, 60] are not candidate interest sequences obtained by selecting consecutive candidate interests in strict sorted order and can therefore be excluded from candidate interest sequences.

[0091] In practical applications, by selecting a preset number of consecutive candidate interests from the sorted candidate interests in the sorted order to determine the candidate interest sequence, the number of determined candidate interest sequences can be reduced, which is conducive to alleviating the workload of the conversion rate prediction model. The conversion rate prediction model does not need to predict every possible interest sequence, which saves computing resources and is also conducive to improving the efficiency of determining the target interest sequence and saving time.

[0092] Optionally, selecting the preset number of consecutive candidate interests from the sorted candidate interests in the sorted order to obtain a candidate interest sequence may specifically include:

[0093] A preset sliding window is used to select the preset number of candidate interests from the sorted candidate interests to obtain a candidate interest sequence; the window length of the preset sliding window is the value represented by the preset number, and the step is the length of one candidate interest.

[0094] In practical applications, a preset sliding window can be used to select a preset number of candidate rights from the sorted candidate rights. The window length of the preset sliding window is the value represented by the preset number, and the stepping is the length of one candidate right. Since the stepping is the length of one candidate right, the candidate rights displayed between adjacent windows of the preset sliding window are also adjacent candidate rights after sorting. For example, there are four types of installment plans: 3 installments, 6 installments, 9 installments, and 12 installments. The corresponding preset number is 4. Assuming the candidate rights are credit points, and the credit point values ​​are 10 credit points, 30 credit points, 50 credit points, 80 credit points, 150 credit points, and 200 credit points, these credit point values ​​can be sorted in order, where each credit point value can occupy three characters. Thus, the length of the candidate rights can be three characters. The window length of the preset sliding window is 12 characters, and the stepping is 3 characters. Alternatively, each credit score value can occupy four characters, where three characters represent the score value and one character represents a null character. Thus, the candidate equity can be four characters long, and the preset sliding window length is 16 characters with a step size of 4 characters. The candidate equity sequence determined using the preset sliding window can include [10, 30, 50, 80], [30, 50, 80, 150], [50, 80, 150, 200], and so on.

[0095] In practical applications, the length of candidate stakes can be set according to actual needs, and the length and step of the sliding window can also be set according to actual needs, and are not specifically limited here.

[0096] The conversion rate prediction model in the embodiments of this specification can be obtained based on relevant training data. Figure 2 The method may further include:

[0097] Acquire training data; the training data includes historical business data of historical users whose historical rights and interests information is obtained; the historical rights and interests information is rights and interests information determined based on the plurality of candidate rights and interests sequences; the historical business data includes historical order information of the historical users, the historical rights and interests information, and information on whether the historical users applied for installment services based on the historical rights and interests information;

[0098] The training data is used to train a neural network classification model to obtain the conversion rate prediction model.

[0099] In the embodiments of this specification, training data can be determined based on historical business data of historical users who have obtained historical rights and interests information. The historical users can be users who have been pushed historical rights and interests information in the past, or they can be test users who participate in the test to collect training data, and there is no specific limitation on this.

[0100] In an embodiment of this specification, historical business data may include historical order information and historical rights information of a historical user, and information about whether the historical user applied for installment business based on the historical rights information. If the historical user applied for installment business based on the historical rights information, the historical business data may also include the installment form of the installment business applied for by the historical user. A neural network classification model is trained using training data. The neural network classification model can learn the characteristics of the data through the training process. After training, the neural network classification model can predict the user's conversion rate for each installment form under the conditions of the candidate rights sequence based on the candidate rights sequence and order data, thereby obtaining a conversion rate prediction model.

[0101] In practical applications, the termination conditions for the process of training a neural network classification model using training data may include reaching a convergence condition or reaching a preset iteration threshold, wherein the convergence condition may include the loss value of the model loss function reaching a preset threshold. For example, if the preset iteration threshold is 500 times, the neural network classification model is trained using the training data, and the training ends after 500 iterations. For another example, if the convergence condition is set to the loss value of the model loss function being less than 0.01, the neural network classification model is trained using the training data, and the training ends when the loss value of the model loss function is less than 0.01.

[0102] In practical applications, you can also divide the training data samples into a training set and a test set. After training the model with the samples in the training set, you can use the samples in the test set to test the performance of the trained model (such as the model's accuracy, precision, and recall). When the model's performance meets the requirements, you can end the training. If the performance of the model is found to be substandard after testing, you can adjust the model's training parameters and continue training, or use new samples to further train the model.

[0103] Figure 4 This is a flow chart of a conversion rate prediction model training method provided in the embodiments of this specification. Figure 4 As shown, the conversion rate prediction model training method may specifically include the following steps:

[0104] Step 402: Acquire training data. The training data includes historical business data of historical users whose historical rights and interests information has been acquired; the historical rights and interests information is rights and interests information determined based on the plurality of candidate rights and interests sequences; and the historical business data includes historical order information of the historical users, the historical rights and interests information, and information on whether the historical users applied for installment services based on the historical rights and interests information.

[0105] Step 404: Divide the training data into a training set and a test set. The training data in the training set may be completely different from the training data in the test set, or may be partially the same, which is not specifically limited.

[0106] Step 406: Input the features corresponding to the training samples in the training set into the neural network classification model. During training, the neural network classification model can learn the features of the data in the training samples in the training set. After training, the neural network classification model can predict the user's conversion rate for each installment type under the candidate benefit sequence and order data.

[0107] Step 408: The model performs forward propagation, calculates loss value and performs backward propagation.

[0108] Step 410: Model weight update.

[0109] Step 412: Determine whether the model training termination condition is met. If the model training termination condition is met, jump to step 414; if not, jump back to step 406. The model training termination condition may include: the loss value of the model loss function reaches a preset threshold, or the number of iterations reaches a preset iteration threshold.

[0110] Step 414: Evaluate the accuracy of the neural network classification model using the training samples in the test set.

[0111] Step 416: Determine whether the accuracy of the model meets the standard. If the accuracy of the model meets the standard, jump to step 418; if the accuracy of the model does not meet the standard, jump to step 420.

[0112] Step 418: Obtain the trained conversion rate prediction model.

[0113] Step 420: Adjust model training parameters or update training samples, and jump back to step 406.

[0114] In order to improve the accuracy of the rights and interests push, the embodiments of this specification can also be combined with the user's characteristics to push. Figure 2 The method may further include:

[0115] Obtaining user feature data of the user;

[0116] The method of using a conversion rate prediction model to predict the user's conversion rate for the various installment forms under the conditions of each candidate benefit sequence for a plurality of candidate benefit sequences specifically includes:

[0117] For any candidate benefit sequence among the several candidate benefit sequences, the any candidate benefit sequence, the user characteristic data and the order characteristic data of the order are provided to the conversion rate prediction model, and the conversion rate prediction model is used to predict the conversion rate of the user for each installment form under the conditions of any candidate benefit sequence.

[0118] In the embodiments of this specification, the user characteristic data of the user may include the characteristics of the user itself, such as: the user's credit limit, the number of visits of the user in the past 7 days, the user's occupation information, the user's education level, the user's historical installment information, the user's asset information, the user's transaction habit information, the user's preference for installment channels, etc.

[0119] In the embodiments of this specification, the order feature data of an order may include but is not limited to: order amount, order source, order identification information, order payment method, order payee information, order payee information, etc.

[0120] In actual applications, the user's terminal device can obtain the user's user feature data and, after the user's authorization, can report the user's user feature data to the server. Alternatively, after the user's authorization, the user's terminal device can also report the user's user feature data to the server once every preset time period. Alternatively, after the user's authorization, the server can also obtain the user's user feature data while obtaining the user's order information of the user's pending order from the user's terminal device. There is no specific limitation on this.

[0121] In practical applications, during the model training phase, the training data may include user feature data of the training users. Thus, during the model training process, the model can also learn the user feature information of the training users. After training, the neural network classification model can predict the user's conversion rate for various installment types under the candidate benefit sequence based on the candidate benefit sequence, user feature data, and order data. The training users may be users who have previously received historical benefit information push notifications, or they may be test users participating in testing to collect training data, without specific limitations.

[0122] In an embodiment of the present specification, the user's user feature data is also input into the conversion rate prediction model, so that the conversion rate prediction model can also refer to the user's user feature data during the calculation process, thereby referring to the user's own characteristics, which is conducive to improving the accuracy of the results output by the model, and then conducive to improving the accuracy of rights and interests push, and conducive to improving user conversion rate.

[0123] In the embodiment of this specification, the candidate equity sequence can also be selected in combination with the order amount to further reduce the number of predictions required by the conversion rate prediction model and improve efficiency. Figure 2 The method may further include:

[0124] For any candidate equity sequence among the plurality of candidate equity sequences, determining a first equity value of a candidate equity corresponding to a first installment form among the various installment forms in the candidate equity sequence, and determining a second equity value of a candidate equity corresponding to a second installment form among the various installment forms in the candidate equity sequence; the total installment duration of the first installment form is greater than the total installment duration of the second installment form;

[0125] Determining, based on the amount information, a first installment benefit value corresponding to the first installment form and a second installment benefit value corresponding to the second installment form;

[0126] Determining whether a first difference between the first installment income value and the first equity value is greater than or equal to a second difference between the second installment income value and the second equity value;

[0127] If the first difference is greater than or equal to the second difference, determining any candidate equity sequence as a candidate equity sequence that matches the order information;

[0128] Correspondingly, the use of the conversion rate prediction model to predict the user's conversion rate for the various installment forms under the conditions of each candidate benefit sequence may specifically include:

[0129] The conversion rate prediction model is used to predict the conversion rate of the user for the various installment forms under the conditions of each candidate benefit sequence matching the order information.

[0130] In the embodiments of this specification, after determining the candidate equity sequence, the equity value corresponding to each installment form can be determined. For example, there are 4 types of installment forms, namely: 3 installments, 6 installments, 9 installments, and 12 installments. Assuming that the candidate equity is credit points, and the credit point values ​​are divided into: 10 credit points, 20 credit points, 30 credit points, 40 credit points, 50 credit points, and 60 credit points, and a candidate equity sequence is [20, 30, 40, 50], it can be determined that the equity value corresponding to 3 installments is 20 credit points, the equity value corresponding to 6 installments is 30 credit points, the equity value corresponding to 9 installments is 40 credit points, and the equity value corresponding to 12 installments is 50 credit points.

[0131] In the embodiments of this specification, each installment payment type has a corresponding interest rate. The installment income value corresponding to each installment payment type can be calculated based on the pending amount information in the order information and the interest rate corresponding to each installment payment type. Specifically, the installment income value corresponding to each installment payment type can be calculated by multiplying the pending amount information in the order information by the interest rate corresponding to each installment payment type. Alternatively, other methods can be used to calculate the installment income value corresponding to each installment payment type, and this is not specifically limited.

[0132] In an embodiment of the present specification, the first difference between the first installment income value and the first equity value can represent the net income of the first installment form, the second difference between the second installment income value and the second equity value can represent the net income of the second installment form, the total installment duration of the first installment form is greater than the total installment duration of the second installment form, and when the first difference is greater than or equal to the second difference, any candidate equity sequence is determined as a candidate equity sequence that matches the order information, otherwise any candidate equity sequence is screened out. In this way, some candidate equity sequences that do not meet the income requirements can be screened out. On the one hand, the number of candidate equity sequences that match the order information can be reduced, which is conducive to reducing the workload of the conversion rate prediction model, saving computing resources, and also helping to improve the efficiency of determining the target equity sequence and saving time. On the other hand, it can also protect the income of the installment business issuer or service provider and promote it to better provide services to users.

[0133] Optional, Figure 2 The method may further include:

[0134] For any candidate equity sequence among the plurality of candidate equity sequences, determining an average equity value corresponding to each candidate equity included in the candidate equity sequence;

[0135] Determining whether a quotient of the amount value represented by the amount information and the average equity value is greater than or equal to a preset threshold;

[0136] If the quotient is greater than or equal to a preset threshold, determining any candidate equity sequence as a candidate equity sequence that matches the order information;

[0137] Correspondingly, the use of the conversion rate prediction model to predict the user's conversion rate for the various installment forms under the conditions of each candidate benefit sequence may specifically include:

[0138] The conversion rate prediction model is used to predict the conversion rate of the user for the various installment forms under the conditions of each candidate benefit sequence matching the order information.

[0139] In practical applications, for any candidate equity sequence among several candidate equity sequences, the average equity value corresponding to each candidate equity contained in the candidate equity sequence can be determined by summing the sum of each candidate equity in the candidate equity sequence and dividing it by the number of candidate equity in the candidate equity sequence to obtain the average equity value. For example, for the candidate equity sequence [1, 2, 3, 4], 1+2+3+4=10, and 10 / 4=2.5, thus calculating the average equity value of the candidate equity sequence as 2.5.

[0140] In the embodiments of this specification, the preset threshold can be set and adjusted based on business needs. For example, the preset threshold can be 30, 50, or 80, and this is not specifically limited. For example, assuming the preset threshold is set to 50, the amount value represented by the amount information is 100 yuan, and the average equity value calculated from candidate equity sequence A is 2, then the quotient of the amount value represented by the amount information and the average equity value = 100 / 2 = 50. If this quotient satisfies the condition of being equal to the preset threshold, candidate equity sequence A can be determined as the candidate equity sequence that matches the order information.

[0141] In actual applications, if the quotient of the amount value represented by the amount information and the average equity value of a candidate equity sequence is less than a preset threshold, it may result in a lower net income corresponding to the candidate equity sequence, affecting the marketing income of the installment service provider. Therefore, by judging whether the quotient of the amount value represented by the amount information and the average equity value is greater than or equal to the preset threshold, some candidate equity sequences whose quotient of the amount value and the average equity value is less than the preset threshold can be screened out. On the one hand, the number of candidate equity sequences that match the order information can be reduced, which is conducive to reducing the workload of the conversion rate prediction model, saving computing resources, and improving the efficiency of determining the target equity sequence and saving time. On the other hand, it can also protect the income of the installment service issuer or service provider and promote it to better provide services to users.

[0142] In actual applications, due to some business needs or some restrictions on the issuance of rights and interests, it may be necessary to adjust the candidate rights and interests based on the obtained candidate rights and interests sequence. For example, in the case of 3-installment payment, the user has already enjoyed the interest-free discount. Assuming that the obtained candidate rights and interests sequence represents the rights and interests regarding the immediate reduction amount, according to business rules, the user cannot enjoy multiple benefits superimposed on each other. For the installment form of 3-installment payment, the user no longer enjoys the right and interests of the immediate reduction amount. In the embodiment of this specification, the rights and interests series can be adjusted by setting the rights and interests adjustment matrix, and the rights and interests that match the user are selected from the adjusted rights and interests sequence and pushed to the user. Optional, Figure 2 The method may further include:

[0143] Generate a rights and interests adjustment matrix based on the rights and interests issuance rules corresponding to the installment business; the number of adjustment parameters included in the rights and interests adjustment matrix is ​​the same as the number of types of installment forms, and each installment form corresponds to one adjustment parameter; the adjustment parameter is a parameter used to indicate whether the installment form corresponding to the adjustment parameter is entitled to rights and interests;

[0144] Calculating the product of the equity adjustment matrix and each candidate equity sequence to obtain each adjusted candidate equity sequence;

[0145] Correspondingly, the use of the conversion rate prediction model to predict the user's conversion rate for the various installment forms under the conditions of each candidate benefit sequence may specifically include:

[0146] The conversion rate prediction model is used to predict the conversion rate of the user for the various installment forms under the conditions of the various adjusted candidate benefit sequences.

[0147] In the embodiments of this specification, the rights issuance rules corresponding to the installment business can be the rights issuance rules formulated by the issuer of the installment business according to business needs, such as: those who enjoy interest-free installment forms will no longer enjoy other rights, some specified installment forms cannot enjoy rights, and some installment forms cannot enjoy rights when the specified order amount is less than a certain threshold, etc.

[0148] In the embodiment of this specification, the number of adjustment parameters included in the equity adjustment matrix is ​​the same as the number of types of installment forms, and one installment form corresponds to one adjustment parameter. The equity adjustment matrix can be a binary matrix, which can also be called a logical matrix. It can contain at least one of the elements 1 or 0, or it can be a one-hot vector. Specifically, if a certain installment form can enjoy benefits in the equity issuance rules corresponding to the installment business, the adjustment parameter corresponding to the installment form can be 1; if a certain installment form cannot enjoy benefits in the equity issuance rules corresponding to the installment business, the adjustment parameter corresponding to the installment form can be 0. For example, if there are four installment payment options: 3, 6, 9, and 12 installments, the equity adjustment matrix contains 4 adjustment parameters. If the 3-installment plan is interest-free, then the 3-installment plan will not receive any other benefits, and the corresponding adjustment parameter for the 3-installment plan can be 0. If, in this transaction, 6, 9, and 12 installments all qualify for benefits, the corresponding adjustment parameters for each of the 6, 9, and 12 installments can be 1, resulting in an equity adjustment matrix of [0, 1, 1, 1]. For example, if a candidate equity sequence is [20, 30, 40, 50], multiplying this candidate equity sequence by the equity adjustment matrix yields an adjusted candidate equity sequence of [0, 30, 40, 50]. This adjusted candidate equity sequence indicates that 3 installments earn 0 credit points, 6 installments earn 30 credit points, 9 installments earn 40 credit points, and 12 installments earn 50 credit points.

[0149] In practical applications, based on the equity payment rules corresponding to the installment business, the generated equity adjustment matrix can be one or more. If multiple equity adjustment matrices are generated, they can all be multiplied by the candidate equity sequence to obtain an adjusted candidate equity sequence. For example, based on the equity payment rule "3-installment business enjoys interest-free benefits but no other benefits," the equity adjustment matrix can be generated as: [0, 1, 1, 1]; based on the equity payment rule "Specify that 3 and 12 installments are not eligible for the immediate reduction benefit," the equity adjustment matrix can be generated as: [0, 1, 1, 0]. Assuming a candidate equity sequence is [20, 30, 40, 50], multiplying this candidate equity sequence with the two generated equity adjustment matrices above yields an adjusted candidate equity sequence of [0, 30, 40, 0]. This adjusted candidate equity sequence indicates that 3 installments earn 0 credit points, 6 installments earn 30 credit points, 9 installments earn 40 credit points, and 12 installments earn 0 credit points.

[0150] In the embodiments of the present specification, an equity adjustment matrix is ​​generated according to the equity issuance rules corresponding to the installment business, and the product of the equity adjustment matrix and each candidate equity sequence is calculated to obtain each adjusted candidate equity sequence. When determining the candidate equity sequence that needs to be calculated by the conversion rate prediction model, the equity issuance rules corresponding to the installment business can be fully considered, and only the adjusted candidate equity sequence that meets the equity issuance rules is calculated using the conversion rate prediction model. On the one hand, it can ensure that the determined target equity sequence meets the equity issuance rules corresponding to the installment business. On the other hand, it can also reduce the workload of the conversion rate prediction model, save computing resources, and help improve the efficiency of determining the target equity sequence and save time.

[0151] In actual applications, when considering the user's rights and interests, the service provider's costs must also be considered. In order to improve the experience of both the user and the service provider, the embodiments of this specification can also combine marketing costs to push rights and interests to users. Figure 2 The method may further include:

[0152] Determining the marketing cost of each candidate equity sequence based on the conversion rate;

[0153] Determining, based on the marketing cost, the conversion rate, and a preset cost constraint, a comprehensive conversion rate of the user for each of the various installment forms under the conditions of each candidate benefit sequence;

[0154] Correspondingly, determining a target equity sequence that meets a preset conversion rate condition from the plurality of candidate equity sequences based on the conversion rate may specifically include:

[0155] Based on the comprehensive conversion rate, the candidate equity sequence with the highest comprehensive conversion rate among the plurality of candidate equity sequences is determined as the target equity sequence.

[0156] In the embodiments of this specification, marketing costs may refer to the costs incurred by an enterprise during the marketing process from the initial owner to the final owner of a product, and are a necessary investment in the enterprise's profits. The marketing costs of a candidate equity sequence may refer to the amount required by the equity issuer to push the equity quotas corresponding to the various installment forms of the candidate equity sequence to users. The preset cost constraint may refer to the monetary limit set by the equity issuer that marketing costs must not exceed.

[0157] In actual applications, when issuing equity, issuers of installment businesses can also consider marketing costs and preset cost constraints to avoid marketing costs exceeding the preset cost constraints, and optimize cost control to increase net profits.

[0158] In the embodiments of this specification, the marketing costs for each candidate benefit sequence can be first determined based on the user's conversion rate for various installment plans under the conditions of each candidate benefit sequence. Then, based on the marketing costs, the conversion rates, and a preset cost constraint, the user's comprehensive conversion rates for various installment plans under the conditions of each candidate benefit sequence can be determined. The candidate benefit sequence with the highest comprehensive conversion rate can then be determined as the target benefit sequence. Since the methods for determining the marketing costs and comprehensive conversion rates for each candidate benefit sequence will be explained in detail in subsequent embodiments of this specification, they will not be detailed here.

[0159] Optionally, for any candidate benefit sequence among the multiple candidate benefit sequences, the conversion rate includes an installment conversion rate of the user using each installment form to process the pending order under the conditions of the candidate benefit sequence; determining the marketing cost of each candidate benefit sequence based on the conversion rate may specifically include:

[0160] For any candidate equity sequence, calculate the product of each candidate equity in the candidate equity sequence and the corresponding installment conversion rate;

[0161] The products are added together to obtain the marketing cost of any candidate equity sequence.

[0162] In an embodiment of this specification, for any candidate benefit sequence, a conversion rate prediction model can be used to predict the installment conversion rate of each installment type for a user to process pending orders under the conditions of the candidate benefit sequence. Subsequently, the marketing cost of the candidate benefit sequence can be obtained by summing the products of each candidate benefit in the candidate benefit sequence and the corresponding installment conversion rate. For example, if there are four installment types: 3 installments, 6 installments, 9 installments, and 12 installments, and a candidate benefit sequence is [20, 30, 40, 50], and the conversion rate prediction model predicts that the user's installment conversion rate for 3 installments is 0.2, the installment conversion rate for 6 installments is 0.2, the installment conversion rate for 9 installments is 0.3, and the installment conversion rate for 12 installments is 0.1 under the conditions of the candidate benefit sequence, then the marketing cost of the candidate benefit sequence = 20 × 0.2 + 30 × 0.2 + 40 × 0.3 + 50 × 0.1 = 27.

[0163] Optionally, determining the comprehensive conversion rate of the user for the various installment forms under the conditions of each candidate benefit sequence based on the marketing cost, the conversion rate, and a preset cost constraint may specifically include:

[0164] Based on the marketing cost, the conversion rate and the preset cost constraint, the variable value of the dual variable is calculated using an operations optimization algorithm;

[0165] Based on the variable values, the Lagrange dual algorithm is used to determine the comprehensive conversion rate of the user for the various installment forms under the conditions of each candidate equity sequence.

[0166] In the embodiments of this specification, the operations research optimization algorithm is to analyze and solve the problem by establishing a mathematical model and applying an optimization algorithm within the framework of operations research, so as to achieve the goal of an optimal solution or a solution close to the optimal solution. The operations research optimization algorithm generally includes the following key elements and steps: 1. Decision variables: represent the variables to be optimized. 2. Objective function: represents the function to be maximized or minimized. 3. Constraints: describe the value range and other restrictions of the decision variables. 4. Modeling: abstract the actual problem into a mathematical model. 5. Solve λ: apply the optimization algorithm to solve the model to obtain the optimal solution or an approximate optimal solution.

[0167] Specifically, the comprehensive conversion rate can be calculated using the following formula, combining factors such as marketing costs and preset cost constraints:

[0168] x ij =c ij (1-λ(a j -A))

[0169] Among them, x ij It can represent the comprehensive conversion rate of user i for various installment forms under the jth candidate equity sequence; j can take values ​​from 1 to m, where m is the number of candidate equity sequences; c ij It can represent the conversion rate of user i under the jth candidate equity sequence condition predicted by the conversion rate prediction model; λ can represent the dual variable; a j Can represent the marketing cost of the jth candidate equity sequence; A can represent the preset cost constraint. This formula can be used to calculate the comprehensive conversion rate of users under the conditions of each candidate equity sequence.

[0170] For example: Assume that the variable value of the dual variable calculated using the operations optimization algorithm is 0.031, the marketing cost of the candidate equity sequence [10, 20, 30, 40] is 10.8 credit points, the preset cost constraint is 33 credit points, and the conversion rate prediction model predicts that the user's conversion rate under the candidate equity sequence [10, 20, 30, 40] is 0.09768. Then, the comprehensive conversion rate of the user under the candidate equity sequence [10, 20, 30, 40] = 0.09768 × (1 - 0.031 × (10.8 - 33)) = 0.165.

[0171] In practical applications, after calculating the comprehensive conversion rate of a user under the conditions of each candidate equity sequence, the candidate equity sequence with the highest comprehensive conversion rate corresponding to each candidate equity sequence can be determined as the target equity sequence. Specifically, after calculating the comprehensive conversion rate of the user under the conditions of each candidate equity sequence, the comprehensive conversion rates can be sorted in order from small to large or from large to small, and the candidate equity sequence with the highest comprehensive conversion rate in the first or last position after sorting can be determined as the target equity sequence. Alternatively, based on the comprehensive conversion rates corresponding to each candidate equity sequence, the candidate equity sequence with the highest comprehensive conversion rate can be determined as the target equity sequence by rearranging the calculation formula. Alternatively, other methods can be used to determine the candidate equity sequence with the highest comprehensive conversion rate, and this is not specifically limited.

[0172] Among them, the candidate equity sequence with the highest comprehensive conversion rate can be obtained by using the following rearrangement calculation formula:

[0173]

[0174] Among them, c ij 、a j The specific meaning of A is the same as that of the various parameters in the above formula for calculating the comprehensive conversion rate, which will not be repeated here. This can represent selecting the jth candidate equity sequence with the highest overall conversion rate from m candidate equity sequences. The embodiments of this specification summarize that by rearranging the calculation formula, the jth candidate equity sequence with the highest overall conversion rate under the conditions of each candidate equity sequence can be calculated. The jth candidate equity sequence can then be determined as the target equity sequence, and target equity information determined based on the target equity sequence can be pushed to the user.

[0175] In practical applications, operations research optimization algorithms may include but are not limited to: goal programming algorithms, integer programming algorithms, stochastic programming algorithms, multi-objective optimization algorithms, genetic algorithms, dynamic programming algorithms, linear programming algorithms, nonlinear programming algorithms, etc.

[0176] In the embodiments of this specification, a dual variable is an important concept in the dual problem of linear programming. A dual variable is a variable defined for each constraint in the original problem when constructing the dual problem of a linear programming problem. These variables play a role in the dual problem corresponding to the variables in the original problem, but their values ​​and properties are restricted by the constraints of the dual problem.

[0177] In the embodiments of this specification, the Lagrange dual algorithm is a technique widely used in optimization theory, particularly suitable for solving optimization problems with constraints. Based on the principle of Lagrange duality, the Lagrange dual algorithm can transform the original constrained optimization problem into a dual problem for solution, thereby simplifying the solution of the original problem. By transforming the dual problem into a more easily solvable form, the Lagrange dual algorithm can find the optimal solution or a near-optimal solution to the original problem.

[0178] In practical applications, the variable values ​​of the dual variables are calculated using an operations optimization algorithm based on marketing costs, conversion rates, and preset cost constraints. Based on the variable values, the Lagrange dual algorithm is used to determine the comprehensive conversion rates of users for various installment forms under the conditions of each candidate equity sequence. The comprehensive conversion rates of various installment forms can be determined based on marketing costs, and the candidate equity sequence with the highest comprehensive conversion rate among several candidate equity sequences can be determined as the target equity sequence. This can minimize marketing cost investment while maximizing marketing benefits.

[0179] Figure 5 This is an overall flow chart of a method for pushing benefits provided in the embodiment of this specification. Figure 5 As shown, the method of pushing rights and interests may specifically include the following steps:

[0180] Step 502: Obtain order information of the user's pending order and user feature data.

[0181] Step 504: Determine various installment forms of the installment business corresponding to the amount information to be processed included in the order information.

[0182] Step 506: Sort the candidate rights according to their rights values, and select a preset number of candidate rights from the sorted candidate rights in the sorted order to obtain a plurality of candidate rights sequences.

[0183] It should be noted that, in other embodiments, step 506 may also be completed before step 502 .

[0184] Step 508: Eliminate the candidate equity sequences that do not meet the first condition and the candidate equity sequences that do not meet the second condition from the obtained candidate equity sequences.

[0185] The first condition is that the first difference between the first installment income value and the first equity value is greater than or equal to the second difference between the second installment income value and the second equity value. The first installment income value is the installment income value corresponding to the first installment form determined based on the amount information in the order information; the second installment income value is the installment income value corresponding to the second installment form determined based on the amount information in the order information; the total installment duration of the first installment form is greater than the total installment duration of the second installment form; the first equity value is the equity value of the candidate equity corresponding to the first installment form in any candidate equity sequence; and the second equity value is the equity value of the candidate equity corresponding to the second installment form in any candidate equity sequence.

[0186] The second condition is that the quotient of the amount represented by the pending amount information included in the order information and the average equity value is greater than or equal to a preset threshold. The average equity value is the average equity value corresponding to each candidate equity contained in any candidate equity sequence among the plurality of candidate equity sequences.

[0187] In practical applications, the candidate equity sequences that meet the first and second conditions can be selected from the obtained candidate equity sequences and then used for subsequent steps. Alternatively, the selection can be performed based on either the first or second condition.

[0188] Step 510: Generate an equity adjustment matrix according to the equity issuance rules corresponding to the installment business.

[0189] Step 512: Calculate the product of the equity adjustment matrix and each candidate equity sequence to obtain each adjusted candidate equity sequence.

[0190] Step 514: Utilize the conversion rate prediction model to predict the user's conversion rate for various installment forms under the conditions of each adjusted candidate benefit sequence.

[0191] Step 516: Based on the conversion rate, determine the marketing cost of each candidate benefit sequence, and based on the marketing cost, conversion rate and preset cost constraints, determine the comprehensive conversion rate of the user for various installment forms under the conditions of each candidate benefit sequence.

[0192] Step 518: Based on the comprehensive conversion rate, determine the candidate equity sequence with the highest comprehensive conversion rate among the candidate equity sequences as the target equity sequence.

[0193] Step 520: According to the target equity sequence, the target equity information corresponding to various installment forms is determined, and the target equity information is pushed to the user.

[0194] Figure 5In the method, after obtaining the order information of the user's pending order, the various installment forms of the installment business corresponding to the pending amount information included in the order information can be determined, and the conversion rate prediction model can be used to predict the user's conversion rate for various installment forms under the conditions of each candidate benefit sequence. Further, based on the conversion rate, the target benefit information corresponding to the target benefit sequence that meets the preset conversion rate conditions is determined from several candidate benefit sequences and pushed to the user. Thus, based on the order information of the user's pending order, the model can be used to predict the conversion rate, which can be faster and more accurate, without excessive human intervention, which is conducive to improving the efficiency of target benefit determination and eliminating the need for excessive use of human resources. On the other hand, by limiting the candidate benefit corresponding to the installment form with a longer total installment duration to a higher benefit value, precise delivery of benefits can be achieved, which is conducive to increasing users' willingness to use benefits, improving conversion rates, and thus saving network or server computing resources.

[0195] Based on the same idea, the embodiments of this specification also provide a device corresponding to the above method.

[0196] Figure 6 The embodiments of this specification provide corresponding Figure 2 A schematic diagram of the structure of a device for pushing rights. Figure 6 As shown, the device may include:

[0197] The information acquisition module 602 is used to obtain the order information of the user's pending order; the order information includes the amount information to be processed.

[0198] The first determining module 604 is configured to determine various installment forms of the installment business corresponding to the amount information.

[0199] The model prediction module 606 is used to predict the conversion rate of the user for the various installment forms under the conditions of each candidate benefit sequence using a conversion rate prediction model; the number of candidate benefits included in a candidate benefit sequence is the same as the number of types of installment forms, and each installment form corresponds to one candidate benefit; for any two installment forms among the various installment forms, the candidate benefit corresponding to the installment form with a longer total installment duration has a higher benefit value.

[0200] The second determining module 608 is configured to determine, based on the conversion rate, a target equity sequence that meets a preset conversion rate condition from the plurality of candidate equity sequences.

[0201] The third determining module 610 is configured to determine target equity information corresponding to the various installment forms according to the target equity sequence.

[0202] The information push module 612 is configured to push the target rights and interests information to the user.

[0203] based on Figure 6 The present specification also provides some specific implementation plans of the device, which are described below.

[0204] Optionally, the device may further include:

[0205] The sorting module is used to sort each candidate interest according to the interest value to obtain the sorted candidate interest.

[0206] The candidate equity sequence determination module is used to select a preset number of candidate equity from the sorted candidate equity in the sorted order to obtain a candidate equity sequence; the preset number is equal to the number represented by the number of types in the installment form.

[0207] Optionally, the candidate equity sequence determination module may specifically include:

[0208] The candidate rights sequence determining unit is configured to select a preset number of consecutive candidate rights from the sorted candidate rights in a sorted order to obtain a candidate rights sequence.

[0209] Optionally, the candidate equity sequence determination unit may specifically include:

[0210] The candidate equity sequence determination subunit is configured to use a preset sliding window to select the preset number of candidate equity from the sorted candidate equity to obtain a candidate equity sequence; the window length of the preset sliding window is the value represented by the preset number, and the step is the length of one candidate equity.

[0211] Optionally, the device may further include:

[0212] A training data acquisition module is used to acquire training data; the training data includes historical business data of historical users who have acquired historical rights and interests information; the historical rights and interests information is rights and interests information determined based on the multiple candidate rights and interests sequences; the historical business data includes historical order information of the historical users, the historical rights and interests information, and information on whether the historical users have applied for installment business based on the historical rights and interests information.

[0213] The model training module is used to train the neural network classification model using the training data to obtain the conversion rate prediction model.

[0214] Optionally, the device may further include:

[0215] The fourth determination module is used to determine, for any candidate equity sequence among the several candidate equity sequences, the first equity value of the candidate equity corresponding to the first installment form in each installment form in the any candidate equity sequence, and to determine the second equity value of the candidate equity corresponding to the second installment form in the any candidate equity sequence; the total installment duration of the first installment form is greater than the total installment duration of the second installment form.

[0216] The fifth determining module is used to determine a first installment benefit value corresponding to the first installment form and a second installment benefit value corresponding to the second installment form according to the amount information.

[0217] The first judgment module is used to judge whether a first difference between the first installment income value and the first equity value is greater than or equal to a second difference between the second installment income value and the second equity value.

[0218] A sixth determining module is configured to determine any candidate equity sequence as a candidate equity sequence that matches the order information if the first difference is greater than or equal to the second difference.

[0219] Correspondingly, the model prediction module 606 may specifically include:

[0220] The first prediction unit is configured to use a conversion rate prediction model to predict the user's conversion rate for the various installment forms under the conditions of each candidate benefit sequence matching the order information.

[0221] Optionally, the device may further include:

[0222] The seventh determining module is configured to determine, for any candidate equity sequence among the plurality of candidate equity sequences, an average equity value corresponding to each candidate equity contained in the candidate equity sequence.

[0223] The second judgment module is used to judge whether the quotient of the amount value represented by the amount information and the average equity value is greater than or equal to a preset threshold.

[0224] An eighth determining module is configured to determine any candidate equity sequence as a candidate equity sequence matching the order information if the quotient is greater than or equal to a preset threshold.

[0225] Correspondingly, the model prediction module 606 may specifically include:

[0226] The second prediction unit is configured to use a conversion rate prediction model to predict the user's conversion rate for the various installment forms under the conditions of each candidate benefit sequence matching the order information.

[0227] Optionally, the device may further include:

[0228] The equity adjustment matrix generation module is used to generate an equity adjustment matrix based on the equity issuance rules corresponding to the installment business; the number of adjustment parameters contained in the equity adjustment matrix is ​​the same as the number of types of installment forms, and each installment form corresponds to one adjustment parameter; the adjustment parameter is a parameter used to indicate whether the installment form corresponding to the adjustment parameter has equity.

[0229] The calculation module is configured to calculate the product of the equity adjustment matrix and each candidate equity sequence to obtain each adjusted candidate equity sequence.

[0230] Correspondingly, the model prediction module 606 may specifically include:

[0231] The third prediction unit is configured to use a conversion rate prediction model to predict the user's conversion rate for the various installment forms under the conditions of the various adjusted candidate equity sequences.

[0232] Optionally, the device may further include:

[0233] The user characteristic data acquisition module is used to acquire the user characteristic data of the user.

[0234] Correspondingly, the model prediction module 606 may specifically include:

[0235] The fourth prediction unit is used to provide any candidate benefit sequence among the several candidate benefit sequences, the user characteristic data and the order characteristic data of the order to the conversion rate prediction model, and use the conversion rate prediction model to predict the conversion rate of the user for each installment form under the condition of any candidate benefit sequence.

[0236] Optionally, the device may further include:

[0237] The marketing cost determination module is used to determine the marketing cost of each candidate equity sequence based on the conversion rate.

[0238] The comprehensive conversion rate determination module is used to determine the comprehensive conversion rate of the user for the various installment forms under the conditions of each candidate benefit sequence based on the marketing cost, the conversion rate and a preset cost constraint.

[0239] Correspondingly, the second determining module 608 may specifically include:

[0240] The target equity sequence determining unit is configured to determine, based on the comprehensive conversion rate, a candidate equity sequence with the highest comprehensive conversion rate among the plurality of candidate equity sequences as a target equity sequence.

[0241] Optionally, for any candidate benefit sequence among the plurality of candidate benefit sequences, the conversion rate includes an installment conversion rate of the user using each installment form to process the pending order under the conditions of the candidate benefit sequence; correspondingly, the marketing cost determination module may specifically include:

[0242] The calculation unit is configured to calculate, for any candidate equity sequence, the product of each candidate equity in the candidate equity sequence and the corresponding installment conversion rate.

[0243] The marketing cost determination unit is configured to add up the products to obtain the marketing cost of any candidate equity sequence.

[0244] Optionally, the comprehensive conversion rate determination module may specifically include:

[0245] A calculation unit is used to calculate the variable value of the dual variable using an operations optimization algorithm based on the marketing cost, the conversion rate and a preset cost constraint.

[0246] The comprehensive conversion rate determination unit is used to determine the comprehensive conversion rate of the user for the various installment forms under the conditions of each candidate equity sequence based on the variable value and using the Lagrange dual algorithm.

[0247] Figure 7 This is a schematic diagram of the structure of a device for pushing benefits provided in the embodiment of this specification. Figure 7 As shown, the device 700 may include:

[0248] at least one processor 710; and,

[0249] a memory 730 communicatively connected to the at least one processor;

[0250] The memory 730 stores instructions 720 that can be executed by the at least one processor 710. The instructions are executed by the at least one processor 710 to enable the at least one processor 710 to:

[0251] Obtaining order information of a user's pending order; the order information includes information on the amount to be processed;

[0252] Determine various installment forms of the installment business corresponding to the amount information;

[0253] For several candidate benefit sequences, a conversion rate prediction model is used to predict the user's conversion rate for the various installment forms under the conditions of each candidate benefit sequence. The number of candidate benefits included in a candidate benefit sequence is the same as the number of installment forms, and each installment form corresponds to one candidate benefit. For any two installment forms among the various installment forms, the candidate benefit corresponding to the installment form with a longer total installment duration has a higher equity value.

[0254] Based on the conversion rate, determining a target equity sequence that meets a preset conversion rate condition from the plurality of candidate equity sequences;

[0255] According to the target equity sequence, determining target equity information corresponding to the various installment forms;

[0256] Push the target rights and interests information to the user.

[0257] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments. Figure 7 As for the device shown, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0258] An embodiment of the present specification further provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the above-mentioned method for pushing benefits.

[0259] The above is an illustrative embodiment of a computer-readable storage medium. It should be noted that the technical solution of this storage medium and the technical solution of the aforementioned method for pushing benefits are based on the same concept. For details not described in detail in the technical solution of the storage medium, please refer to the description of the technical solution of the aforementioned method for pushing benefits.

[0260] The foregoing description of this specification describes specific embodiments. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in an order different from that described in the embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order shown or the sequential order to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0261] In the 1990s, technological improvements could be clearly distinguished as either hardware improvements (for example, improvements to circuit structures such as diodes, transistors, and switches) or software improvements (improvements to process flows). However, with the advancement of technology, many process flow improvements today can now be considered direct improvements to hardware circuit structures. Designers almost always create the corresponding hardware circuit structure by programming the improved process flow into the hardware circuit. Therefore, it cannot be said that a process flow improvement cannot be implemented using hardware modules. For example, a programmable logic device (PLD), such as a field programmable gate array (FPGA), is an integrated circuit whose logical function is determined by user programming. Designers can "integrate" a digital system on a PLD by programming it themselves, without having to hire a chip manufacturer to design and produce a dedicated integrated circuit chip. Moreover, nowadays, instead of manually fabricating integrated circuit chips, this programming is mostly done using "logic compiler" software. This is similar to the software compiler used when developing programs. Before compilation, the original code must also be written in a specific programming language, called a hardware description language (HDL). There is not just one HDL, but many, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, RHDL (Ruby Hardware Description Language), etc. The most commonly used ones are VHDL (Very-High-Speed ​​Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art will also understand that by simply programming the method flow in one of these hardware description languages ​​and then programming it into an integrated circuit, a hardware circuit that implements the logic method flow can be easily obtained.

[0262] The controller can be implemented in any suitable manner. For example, the controller can take the form of a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro)processor, logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers, and embedded microcontrollers. Examples of controllers include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320. The memory controller can also be implemented as part of the control logic of the memory. Those skilled in the art will also know that in addition to implementing the controller in a purely computer-readable program code format, the controller can be implemented in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, such a controller can be considered a hardware component, and the devices included therein for implementing various functions can also be considered as structures within the hardware component. Or even, the devices for implementing various functions can be considered as both software modules that implement the method and structures within the hardware component.

[0263] The systems, devices, modules, or units described in the above embodiments may be implemented by computer chips or entities, or by products having certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0264] For the convenience of description, the above devices are described as being divided into various units according to their functions. Of course, when implementing this specification, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0265] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0266] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0267] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0268] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0269] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0270] Memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0271] Computer-readable media includes permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. The information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0272] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0273] This specification may be described in the general context of computer-executable instructions, such as program modules, executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, and the like that perform specific tasks or implement specific abstract data types. This specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media, including storage devices.

[0274] The foregoing is merely an example of the present invention and is not intended to limit the present invention. Various modifications and variations are possible for those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be included within the scope of the claims of the present invention.

Claims

1. A method for pushing benefits, comprising: Get the order information of the user's pending orders; The order information includes the amount to be processed; Determine various installment forms of the installment business corresponding to the amount information; For several candidate benefit sequences, a conversion rate prediction model is used to predict the user's conversion rate for the various installment forms under the conditions of each candidate benefit sequence. The number of candidate benefits included in a candidate benefit sequence is the same as the number of installment forms, and each installment form corresponds to one candidate benefit. For any two installment forms among the various installment forms, the candidate benefit corresponding to the installment form with a longer total installment duration has a higher equity value. Based on the conversion rate, determining a target equity sequence that meets a preset conversion rate condition from the plurality of candidate equity sequences; According to the target equity sequence, determining target equity information corresponding to the various installment forms; Push the target rights and interests information to the user.

2. The method according to claim 1, further comprising: Sort each candidate interest by its value to obtain sorted candidate interests; A preset number of candidate equities are selected from the sorted candidate equities in the sorting order to obtain a candidate equities sequence; the preset number is equal to the number represented by the number of types in the installment form.

3. The method according to claim 2, wherein selecting a preset number of candidate interests from the sorted candidate interests in order of sorting to obtain the candidate interest sequence specifically comprises: The preset number of consecutive candidate rights are selected from the sorted candidate rights in the sorted order to obtain a candidate rights sequence.

4. The method according to claim 3, wherein the step of selecting the preset number of consecutive candidate interests from the sorted candidate interests in the sorted order to obtain a candidate interest sequence comprises: Using a preset sliding window, selecting the preset number of candidate interests from the sorted candidate interests to obtain a candidate interest sequence; The window length of the preset sliding window is the value represented by the preset number, and the step is the length of one candidate equity.

5. The method of claim 1, further comprising: Get training data; The training data includes historical business data of historical users who have acquired historical rights and interests information; the historical rights and interests information is rights and interests information determined based on the plurality of candidate rights and interests sequences; the historical business data includes historical order information of the historical users, the historical rights and interests information, and information on whether the historical users applied for installment services based on the historical rights and interests information; The training data is used to train a neural network classification model to obtain the conversion rate prediction model.

6. The method of claim 1, further comprising: For any candidate equity sequence among the plurality of candidate equity sequences, determining a first equity value of a candidate equity corresponding to a first installment form among the various installment forms in the candidate equity sequence, and determining a second equity value of a candidate equity corresponding to a second installment form among the various installment forms in the candidate equity sequence; The total duration of the first installment form is greater than the total duration of the second installment form; Determining, based on the amount information, a first installment benefit value corresponding to the first installment form and a second installment benefit value corresponding to the second installment form; Determining whether a first difference between the first installment income value and the first equity value is greater than or equal to a second difference between the second installment income value and the second equity value; If the first difference is greater than or equal to the second difference, determining any candidate equity sequence as a candidate equity sequence that matches the order information; The conversion rate prediction model is used to predict the user's conversion rate for each installment form under the conditions of each candidate benefit sequence, specifically including: The conversion rate prediction model is used to predict the conversion rate of the user for the various installment forms under the conditions of each candidate benefit sequence matching the order information.

7. The method of claim 1, further comprising: For any candidate equity sequence among the plurality of candidate equity sequences, determining an average equity value corresponding to each candidate equity included in the candidate equity sequence; Determining whether a quotient of the amount value represented by the amount information and the average equity value is greater than or equal to a preset threshold; If the quotient is greater than or equal to a preset threshold, determining any candidate equity sequence as a candidate equity sequence that matches the order information; The conversion rate prediction model is used to predict the user's conversion rate for each installment form under the conditions of each candidate benefit sequence, specifically including: The conversion rate prediction model is used to predict the conversion rate of the user for the various installment forms under the conditions of each candidate benefit sequence matching the order information.

8. The method of claim 1, further comprising: Generate an equity adjustment matrix based on the equity issuance rules corresponding to the installment business; The number of adjustment parameters included in the equity adjustment matrix is ​​the same as the number of types of installment forms, and each installment form corresponds to one adjustment parameter; the adjustment parameter is a parameter used to indicate whether the installment form corresponding to the adjustment parameter has equity; Calculating the product of the equity adjustment matrix and each candidate equity sequence to obtain each adjusted candidate equity sequence; The conversion rate prediction model is used to predict the user's conversion rate for each installment form under the conditions of each candidate benefit sequence, specifically including: The conversion rate prediction model is used to predict the conversion rate of the user for the various installment forms under the conditions of the various adjusted candidate benefit sequences.

9. The method of claim 1, further comprising: Obtaining user feature data of the user; The method of using a conversion rate prediction model to predict the user's conversion rate for the various installment forms under the conditions of each candidate benefit sequence for a plurality of candidate benefit sequences specifically includes: For any candidate benefit sequence among the several candidate benefit sequences, the any candidate benefit sequence, the user characteristic data and the order characteristic data of the order are provided to the conversion rate prediction model, and the conversion rate prediction model is used to predict the conversion rate of the user for each installment form under the conditions of any candidate benefit sequence.

10. The method of claim 1, further comprising: Determining the marketing cost of each candidate equity sequence based on the conversion rate; Determining, based on the marketing cost, the conversion rate, and a preset cost constraint, a comprehensive conversion rate of the user for each of the various installment forms under the conditions of each candidate benefit sequence; The step of determining a target equity sequence that meets a preset conversion rate condition from the plurality of candidate equity sequences based on the conversion rate specifically includes: Based on the comprehensive conversion rate, the candidate equity sequence with the highest comprehensive conversion rate among the plurality of candidate equity sequences is determined as the target equity sequence.

11. The method according to claim 10, wherein for any candidate benefit sequence among the plurality of candidate benefit sequences, the conversion rate comprises an installment conversion rate of the user using each installment form to process the pending order under the conditions of the candidate benefit sequence; Determining the marketing cost of each candidate equity sequence based on the conversion rate specifically includes: For any candidate equity sequence, calculate the product of each candidate equity in the candidate equity sequence and the corresponding installment conversion rate; The products are added together to obtain the marketing cost of any candidate equity sequence.

12. The method of claim 10, wherein determining the user's comprehensive conversion rate for each installment plan under each candidate benefit sequence based on the marketing cost, the conversion rate, and a preset cost constraint comprises: Based on the marketing cost, the conversion rate and the preset cost constraint, the variable value of the dual variable is calculated using an operations optimization algorithm; Based on the variable values, the Lagrange dual algorithm is used to determine the comprehensive conversion rate of the user for the various installment forms under the conditions of each candidate equity sequence.

13. A device for pushing benefits, comprising: The information acquisition module is used to obtain the order information of the user's pending orders; The order information includes the amount to be processed; A first determining module is used to determine various installment forms of the installment business corresponding to the amount information; A model prediction module is configured to use a conversion rate prediction model to predict, for a plurality of candidate benefit sequences, the user's conversion rate for the various installment forms under the conditions of each candidate benefit sequence; the number of candidate benefits included in a candidate benefit sequence is equal to the number of types of installment forms, and each installment form corresponds to one candidate benefit; for any two installment forms among the various installment forms, the candidate benefit corresponding to the installment form with a longer total installment duration has a higher equity value; A second determining module is configured to determine, based on the conversion rate, a target equity sequence that meets a preset conversion rate condition from the plurality of candidate equity sequences; A third determining module is configured to determine target equity information corresponding to the various installment forms according to the target equity sequence; The information push module is used to push the target rights and interests information to the user.

14. A device for pushing benefits, comprising: at least one processor; as well as, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to: Obtaining order information of a user's pending order; the order information includes information on the amount to be processed; Determine various installment forms of the installment business corresponding to the amount information; For several candidate benefit sequences, a conversion rate prediction model is used to predict the user's conversion rate for the various installment forms under the conditions of each candidate benefit sequence. The number of candidate benefits included in a candidate benefit sequence is the same as the number of installment forms, and each installment form corresponds to one candidate benefit. For any two installment forms among the various installment forms, the candidate benefit corresponding to the installment form with a longer total installment duration has a higher equity value. Based on the conversion rate, determining a target equity sequence that meets a preset conversion rate condition from the plurality of candidate equity sequences; According to the target equity sequence, determining target equity information corresponding to the various installment forms; Push the target rights and interests information to the user.

15. A computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions, when executed by a processor, implement the method for pushing benefits according to any one of claims 1 to 12.

Images