Resource distribution method and device based on causal model, computing equipment and medium

Abstract

The embodiment of the invention discloses a resource distribution method and device based on a causal model, computing equipment and a medium, and the method comprises the steps: predicting a purchase amount interval of an order generated by a target user in the future according to the historical behavior data of the target user; generating a plurality of resources in the purchase amount interval, and predicting the use probability of each resource; predicting an evaluation index residual error corresponding to the target user under each resource by using a causal model; and formulating a resource distribution strategy for the target user according to the use probability of each resource and the corresponding evaluation index residual under each resource. According to the invention, the causal model is introduced into the formulation scene of the resource distribution strategy, the causal relationship of user behaviors can be effectively captured, and the appropriate resource distribution strategy can be automatically formulated for the user in combination with the use probability of each resource and the corresponding evaluation index residual under each resource, so that personalized resource distribution is realized, and the user experience is improved. And the resource distribution effect is improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of Internet technology, and specifically to a resource distribution method, apparatus, computing device and medium based on a causal model. Background Art

[0002] Platforms and merchants often hold various activities to attract users and distribute resources such as coupons to users to attract users to actively place orders. In the prior art, resources are usually distributed to users based on simple rules or statistical models. However, this resource distribution method has certain limitations. It fails to fully consider the causal relationship behind user behavior, fails to effectively capture the complexity of causal relationships, and is difficult to accurately predict the impact of resources on user behavior, resulting in poor resource distribution effects and waste of resources. Summary of the invention

[0003] In view of the above problems, the present application is proposed to provide a method, apparatus, computing device and medium for allocating resources based on a causal model that overcomes the above problems or at least partially solves the above problems.

[0004] According to one aspect of an embodiment of the present application, a method for allocating resources based on a causal model is provided, comprising:

[0005] Obtain the historical behavior data of the target user, and predict the purchase amount range of the target user's future orders based on the historical behavior data;

[0006] Generate multiple resources within the purchase amount range and predict the usage probability of each resource;

[0007] Use the causal model to predict the residual of the evaluation index corresponding to each resource for the target user;

[0008] According to the usage probability of each resource and the corresponding evaluation indicator residual under each resource, a resource distribution strategy is formulated for the target users.

[0009] Furthermore, based on the historical behavior data, the purchase amount range of the target user's future orders is predicted to further include:

[0010] Preprocess and extract features from historical behavior data to determine corresponding correlation features;

[0011] The associated features are input into the amount prediction model, and the amount prediction model predicts the purchase amount range of the target user's future orders based on the associated features.

[0012] Furthermore, the associated features include: user features of the target user, features of the commodity with which the target user has an interactive relationship, and cross-features between the target user and the commodity with which the target user has an interactive relationship.

[0013] Furthermore, generating multiple resources within the purchase amount range and predicting the usage probability of each resource further includes:

[0014] Generate multiple resources in steps within the purchase amount range;

[0015] The associated features are input into the probability prediction model, which predicts the probability of target users using each resource in the future based on the associated features.

[0016] Furthermore, the process of building a causal model includes:

[0017] Obtain the distribution data of each historical resource in the historical activities and the historical behavior data of each user in the historical activities;

[0018] Divide each historical resource into multiple buckets, and build a causal model training data set based on the distribution data and each user's historical behavior data;

[0019] According to the causal model training data set, the residuals of the evaluation indicators corresponding to different users in each bucket are analyzed, and causal modeling is performed to obtain a causal model.

[0020] Furthermore, using the causal model, the residual of the evaluation index corresponding to the target user under each resource is predicted to further include:

[0021] For each resource, determine the bucket to which the resource belongs, the similar users of the target user, and the residual of the evaluation index corresponding to the similar users in the bucket based on the causal model;

[0022] According to the residual of the evaluation index corresponding to the similar users in the bucket and the similarity between the target user and the similar users, the residual of the evaluation index corresponding to the target user in the resource is predicted.

[0023] According to another aspect of an embodiment of the present application, a resource distribution device based on a causal model is provided, including:

[0024] The first prediction module is adapted to obtain historical behavior data of the target user and predict the purchase amount range of the orders generated by the target user in the future based on the historical behavior data;

[0025] The second prediction module is adapted to generate multiple resources within a purchase amount range and predict the use probability of each resource;

[0026] The third prediction module is suitable for using the causal model to predict the residual of the evaluation index corresponding to the target user under each resource;

[0027] The formulation module is suitable for formulating resource distribution strategies for target users based on the usage probability of each resource and the residual of the corresponding evaluation index under each resource.

[0028] According to another aspect of the embodiment of the present application, there is provided a computing device, including: a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus;

[0029] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to execute operations corresponding to the resource distribution method based on the causal model.

[0030] According to another aspect of the embodiments of the present application, a computer storage medium is provided, in which at least one executable instruction is stored, and the executable instruction enables a processor to perform operations corresponding to the resource distribution method based on the causal model as described above.

[0031] According to another aspect of the embodiments of the present application, a computer program product is provided, comprising at least one executable instruction, wherein the executable instruction enables a processor to execute operations corresponding to the resource release method based on the causal model as described above.

[0032] According to the technical solution provided in the embodiment of the present application, by analyzing the user's historical behavior data, the purchase amount range of the order generated in the future is predicted, and then multiple resources are generated within the purchase amount range, thereby realizing personalized resource generation and predicting the usage probability of each resource; introducing the causal model into the resource issuance strategy formulation scenario can effectively capture the causal relationship of user behavior, and conveniently and accurately predict the evaluation index residual corresponding to the user under each resource; according to specific business needs, combined with the usage probability of each resource and the corresponding evaluation index residual under each resource, it can automatically formulate a suitable resource issuance strategy for the user, thereby realizing personalized resource issuance, optimizing the formulation method of resource issuance strategy, better meeting user needs, optimizing user experience, helping to improve the resource issuance effect, effectively increasing the probability of resource use, and reducing resource waste; in addition, it can also be combined with a real-time feedback mechanism to flexibly adjust the resource issuance strategy according to different business needs.

[0033] The above description is only an overview of the technical solution of the embodiment of the present application. In order to more clearly understand the technical means of the embodiment of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the embodiment of the present application more obvious and easy to understand, the specific implementation method of the embodiment of the present application is listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only used for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the embodiments of the present application. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0035] Figure 1 A schematic diagram of a process of a resource distribution method based on a causal model according to an embodiment of the present application is shown;

[0036] Figure 2 A structural block diagram of a resource distribution device based on a causal model according to an embodiment of the present application is shown;

[0037] Figure 3 A schematic diagram of the structure of a computing device according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0038] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0039] Figure 1 A schematic diagram of a resource distribution method based on a causal model according to an embodiment of the present application is shown. Figure 1 As shown, the method comprises the following steps:

[0040] Step S101, obtaining historical behavior data of a target user, and predicting a purchase amount range of orders that the target user will generate in the future based on the historical behavior data.

[0041] In this application, users who need to formulate resource distribution strategies are referred to as target users, where resources may include coupons in the form of full-reduction coupons, discount coupons, etc. In order to accurately formulate resource distribution strategies for target users and improve the effect of resource distribution, it is necessary to obtain the historical behavior data of target users, where the historical behavior data of target users may include user attribute data, interaction data between target users and various commodities on the platform, and commodity data with which target users have an interactive relationship. Specifically, user attribute data may include user identification, user age, user location, user gender and other data; interaction data may include target users' browsing data, purchase data, return order data, etc. for commodities; commodity data includes commodity identification, commodity pictures, commodity price range, commodity type, commodity shipping place and other data.

[0042] Collect historical behavior data of target users, and perform preprocessing and feature extraction on the historical behavior data to determine the associated features corresponding to the target users. Preprocessing refers to a series of processing steps performed on raw data before data analysis or machine learning, so as to convert the raw data into a form suitable for analysis and modeling, so as to improve the quality and applicability of the data and reduce noise. Among them, preprocessing may include data cleaning, data conversion, feature selection, etc. Specifically, data cleaning may include removing duplicate data, filling or deleting missing values, processing outliers, etc. Data conversion may include standardization processing (such as scaling data to the same scale), normalization processing (such as converting data to intervals such as [0, 1] or [-1, 1]), encoding processing (such as converting categorical variables into numerical form through one-hot encoding, label encoding, etc.), etc.

[0043] The pre-processed historical behavior data is subjected to feature extraction to obtain the corresponding correlation features. The correlation features may include: user features of the target user, features of the product with which the target user interacts, and cross-features between the target user and the product with which the target user interacts. Among them, cross-features help reflect the correlation between the user and the product.

[0044] In the present application, an amount prediction model is used to predict the purchase amount range of the orders generated by the target user in the future. The amount prediction model is obtained by training with the historical behavior data of multiple users, and can be used to accurately predict the purchase amount range of the orders generated by the user in the future. Specifically, the historical behavior data of multiple users can be preprocessed and feature extracted to determine the corresponding associated features, and the associated features corresponding to the multiple users are used to train the model to obtain the amount prediction model. In step S101, the associated features corresponding to the target user are input into the amount prediction model, and the amount prediction model predicts the purchase amount range of the orders generated by the target user in the future based on the associated features.

[0045] Step S102, generating multiple resources within the purchase amount range, and predicting the usage probability of each resource.

[0046] After predicting the purchase amount range of the target user's future orders, multiple resources can be generated in the purchase amount range according to the step length. Specifically, multiple resources can be generated in the purchase amount range according to the step length according to different generation factors. In specific applications, the amount reduction ratio corresponding to the resource itself can be used as the generation factor.

[0047] Assuming that the predicted purchase amount range is 50 to 70, and the step length is 10, then by moving within the purchase amount range with a step length of 10, resources with different generation factors can be generated at 50, 60, and 70. For example, the resources with generation factors of 0.1 and 0.2 generated at 50 may include a coupon for a 5 yuan discount on purchases over 50 yuan and a coupon for a 10 yuan discount on purchases over 50 yuan; and so on, multiple resources are generated.

[0048] After generating multiple resources, the associated features corresponding to the target user are input into the probability prediction model, and the probability prediction model predicts the target user's probability of using each resource in the future based on the associated features. In this application, a probability prediction model is used to predict the target user's probability of using each resource in the future. Among them, the probability prediction model is obtained by training with the historical behavior data of multiple users, and can be used to accurately predict the user's probability of using each resource in the future. Specifically, the historical behavior data of multiple users can be preprocessed and feature extracted to determine the corresponding associated features, and the model can be trained using the associated features corresponding to multiple users to obtain a probability prediction model.

[0049] Step S103: using the causal model, predict the residual of the evaluation index corresponding to each resource of the target user.

[0050] In order to fully consider the causal relationship of user behavior in resource distribution, this application constructs a causal model and introduces the causal model into the scenario of resource distribution strategy formulation. Analyze the user's historical behavior data, capture the causal relationship, and construct a causal relationship diagram of user behavior to achieve dynamic modeling of users. Specifically, machine learning algorithms can be used to analyze the historical behavior data of each user, identify the key factors that affect purchase decisions, and use causal inference technology to evaluate the relationship between different factors to obtain a causal model.

[0051] The construction process of the causal model includes: obtaining the issuance data of each historical resource in historical activities and the historical behavior data of each user in historical activities, wherein the historical behavior data of each user includes real historical behavior data and constructed historical behavior data; dividing each historical resource into multiple buckets, and constructing a causal model training data set based on the issuance data and the historical behavior data of each user; according to the causal model training data set, analyzing the residuals of the evaluation indicators corresponding to different users in each bucket, performing causal modeling, and obtaining a causal model.

[0052] Specifically, each historical resource can be divided into multiple buckets according to the generation factor, business scenario, resource type, etc. of the historical resource, and each bucket corresponds to a different intervention strategy, which may include a resource display strategy, etc. The present application realizes effective management of resources in the form of buckets. The constructed causal model records the association between users, buckets, and evaluation index residuals. The evaluation index residual corresponding to a certain user under a certain bucket refers to the residual between the evaluation index data generated by issuing the resources of the bucket to the user and the evaluation index data generated by not issuing the resources of the bucket to the user. The evaluation index data is the indicator data used to evaluate the business situation, for example, it may include order amount, purchase rate, net profit, etc.

[0053] In step S103, for each resource generated in step S102, the bucket to which the resource belongs, the similar users of the target user, and the corresponding evaluation index residuals of the similar users under the bucket are determined according to the causal model; based on the corresponding evaluation index residuals of the similar users under the bucket and the similarity between the target user and the similar users, the corresponding evaluation index residuals of the target user under the resource are predicted.

[0054] Specifically, the bucket to which the resource belongs can be determined based on the resource generation factor, business scenario, resource type, etc.; the cosine similarity algorithm, etc. is used to calculate the similarity between the associated features corresponding to the target user and the associated features corresponding to each user, and the user with the highest similarity is selected as the similar user of the target user, and then the evaluation index residual corresponding to the similar user under the bucket is determined based on the association relationship between the user, bucket, and evaluation index residual recorded in the causal model. Considering that the target user and similar users are relatively similar but still have certain differences, in order to achieve accurate prediction, the evaluation index residual corresponding to the similar user under the bucket and the similarity between the target user and the similar user can be calculated accordingly, and the calculation result is used as the predicted evaluation index residual corresponding to the target user under the resource.

[0055] Step S104: formulate a resource distribution strategy for the target user according to the usage probability of each resource and the residual of the evaluation indicator corresponding to each resource.

[0056] According to specific business needs, combined with the usage probability of each resource and the residual of the corresponding evaluation index under each resource, it can determine which resources among the generated multiple resources are issued to the target user and which resources are not issued to the target user, and automatically formulate appropriate resource issuance strategies for the target user, thereby realizing personalized resource issuance, optimizing the formulation method of resource issuance strategies, better meeting user needs, helping to improve resource issuance effects, and reducing resource waste. The resource issuance strategy includes which resources are issued to the target user.

[0057] In actual applications, an effective feedback mechanism can also be built to adjust resource distribution strategies in a timely and dynamic manner to adapt to changes in business needs. Specifically, the resource distribution process can be continuously optimized by combining the real-time feedback mechanism. In addition, during the implementation process, different user historical behavior data and causal models can be selected according to different business needs to achieve a higher distribution effect.

[0058] The resource distribution method based on the causal model provided in this application was applied to a specific practical scenario, and it was found that the resource distribution effect was significantly improved. Through the formulation of personalized resource distribution strategies, the user's order amount on the platform increased by 10%, the user's purchase rate increased by 20%, and the probability of resource use increased by 25%.

[0059] It should be noted that the information and data involved in the embodiments of the present application (including but not limited to historical behavior data, issuance data, etc.) 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 relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0060] According to the resource issuance method based on the causal model provided in the embodiment of the present application, by analyzing the user's historical behavior data, the purchase amount range of the order generated in the future is predicted, and then multiple resources are generated within the purchase amount range, thereby realizing personalized resource generation and predicting the usage probability of each resource; introducing the causal model into the resource issuance strategy formulation scenario can effectively capture the causal relationship of user behavior, and conveniently and accurately predict the evaluation index residual corresponding to the user under each resource; according to specific business needs, combined with the usage probability of each resource and the corresponding evaluation index residual under each resource, it can automatically formulate a suitable resource issuance strategy for the user, thereby realizing personalized resource issuance, optimizing the formulation method of resource issuance strategy, better meeting user needs, optimizing user experience, helping to improve the resource issuance effect, effectively increasing the probability of resource use, and reducing resource waste; in addition, it can also be combined with a real-time feedback mechanism to flexibly adjust the resource issuance strategy according to different business needs.

[0061] Figure 2 FIG. 4 shows a structural block diagram of a resource distribution device based on a causal model according to an embodiment of the present application. Figure 2 As shown, the device includes: a first prediction module 210, a second prediction module 220, a third prediction module 230 and a formulation module 240.

[0062] The first prediction module 210 is adapted to: obtain historical behavior data of a target user, and predict a purchase amount range of orders generated by the target user in the future based on the historical behavior data.

[0063] The second prediction module 220 is adapted to generate a plurality of resources within a purchase amount range and predict the usage probability of each resource.

[0064] The third prediction module 230 is adapted to: use a causal model to predict the residual of the evaluation index corresponding to the target user under each resource.

[0065] The formulation module 240 is adapted to formulate a resource distribution strategy for a target user according to the usage probability of each resource and the residual of the evaluation indicator corresponding to each resource.

[0066] Optionally, the first prediction module 210 is further adapted to: preprocess and extract features of historical behavior data to determine corresponding associated features; input the associated features into an amount prediction model, and the amount prediction model predicts the purchase amount range of orders generated by the target user in the future based on the associated features.

[0067] Optionally, the associated features include: user features of the target user, features of the commodity with which the target user has an interactive relationship, and cross-features between the target user and the commodity with which the target user has an interactive relationship.

[0068] Optionally, the second prediction module 220 is further adapted to: generate multiple resources in a purchase amount interval according to a step size; input the associated features into a probability prediction model, and the probability prediction model predicts the target user's probability of using each resource in the future based on the associated features.

[0069] Optionally, the process of constructing the causal model includes: obtaining the issuance data of each historical resource in historical activities and the historical behavior data of each user in historical activities; dividing each historical resource into multiple buckets, and constructing a causal model training data set based on the issuance data and the historical behavior data of each user; based on the causal model training data set, analyzing the residuals of evaluation indicators corresponding to different users in each bucket, performing causal modeling, and obtaining a causal model.

[0070] Optionally, the third prediction module 230 is further adapted to: for each resource, determine, according to the causal model, the bucket to which the resource belongs, the similar users of the target user, and the evaluation index residuals corresponding to the similar users under the bucket; based on the evaluation index residuals corresponding to the similar users under the bucket and the similarity between the target user and the similar users, predict the evaluation index residuals corresponding to the target user under the resource.

[0071] The description of each module above refers to the corresponding description in the method embodiment and will not be repeated here.

[0072] According to the resource distribution device based on the causal model provided in the embodiment of the present application, by analyzing the user's historical behavior data, the purchase amount range of the order generated in the future is predicted, and then multiple resources are generated within the purchase amount range, thereby realizing personalized resource generation and predicting the usage probability of each resource; introducing the causal model into the resource distribution strategy formulation scenario can effectively capture the causal relationship of user behavior, and conveniently and accurately predict the evaluation index residual corresponding to the user under each resource; according to specific business needs, combined with the usage probability of each resource and the corresponding evaluation index residual under each resource, it can automatically formulate a suitable resource distribution strategy for the user, thereby realizing personalized resource distribution, optimizing the formulation method of resource distribution strategy, better meeting user needs, optimizing user experience, helping to improve the resource distribution effect, effectively increasing the probability of resource use, and reducing resource waste; in addition, it can also be combined with a real-time feedback mechanism to flexibly adjust the resource distribution strategy according to different business needs.

[0073] An embodiment of the present application provides a non-volatile computer storage medium, which stores at least one executable instruction or computer program, and the executable instruction or computer program can enable a processor to perform operations corresponding to the resource distribution method based on a causal model in any of the above-mentioned method embodiments.

[0074] An embodiment of the present application provides a computer program product, which includes at least one executable instruction or computer program, and the executable instruction or computer program can enable a processor to perform operations corresponding to the resource issuance method based on a causal model in any of the above-mentioned method embodiments.

[0075] Figure 3 A schematic diagram of the structure of a computing device according to an embodiment of the present application is shown. The specific embodiment of the present application does not limit the specific implementation of the computing device.

[0076] like Figure 3 As shown, the computing device may include: a processor (processor) 302 , a communications interface (Communications Interface) 304 , a memory (memory) 306 , and a communication bus 308 .

[0077] The processor 302, the communication interface 304, and the memory 306 communicate with each other via the communication bus 308. The communication interface 304 is used to communicate with other devices such as a client or other server network elements. The processor 302 is used to execute the program 310, which can specifically execute the relevant steps in the above-mentioned embodiment of the resource distribution method based on the causal model for computing devices.

[0078] Specifically, the program 310 may include program codes, which include computer operation instructions.

[0079] The processor 302 may be a central processing unit (CPU), or an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application. The one or more processors included in the computing device may be processors of the same type, such as one or more CPUs; or processors of different types, such as one or more CPUs and one or more ASICs.

[0080] The memory 306 is used to store the program 310. The memory 306 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0081] Program 310 can be specifically used to enable processor 302 to execute the resource distribution method based on the causal model in any of the above-mentioned method embodiments. The specific implementation of each step in program 310 can refer to the corresponding description in the corresponding steps and units in the above-mentioned resource distribution embodiment based on the causal model, which will not be repeated here. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process of the above-described devices and modules can refer to the corresponding process description in the above-mentioned method embodiment, which will not be repeated here.

[0082] The algorithm and display provided herein are not inherently related to any specific computer, virtual system or other equipment. Various general purpose systems can also be used together with the teachings based on this. According to the above description, it is obvious to construct the structure required for this type of system. In addition, the present application embodiment is not directed to any specific programming language yet. It should be understood that various programming languages ​​can be utilized to realize the content of the present application embodiment described herein, and the above description of specific languages ​​is to disclose the best mode of implementation of the present application embodiment.

[0083] In the description provided herein, a large number of specific details are described. However, it is understood that the embodiments of the present application can be practiced without these specific details. In some instances, well-known methods, structures and techniques are not shown in detail so as not to obscure the understanding of this description.

[0084] Similarly, it should be understood that in order to streamline the present disclosure and help understand one or more of the various inventive aspects, in the above description of the exemplary embodiments of the present application, the various features of the embodiments of the present application are sometimes grouped together into a single embodiment, figure, or description thereof. However, the disclosed method should not be interpreted as reflecting the following intention: the claimed embodiments of the present application require more features than the features clearly stated in each claim. More specifically, as reflected in the claims below, the inventive aspects are less than all the features of the single embodiment disclosed above. Therefore, the claims following the specific embodiment are hereby explicitly incorporated into the specific embodiment, wherein each claim itself serves as a separate embodiment of the embodiment of the present application.

[0085] Those skilled in the art will appreciate that the modules in the devices in the embodiments may be adaptively changed and arranged in one or more devices different from the embodiments. The modules or units or components in the embodiments may be combined into one module or unit or component, and in addition they may be divided into a plurality of submodules or subunits or subcomponents. Except that at least some of such features and / or processes or units are mutually exclusive, all features disclosed in this specification (including the accompanying claims, abstracts and drawings) and all processes or units of any method or device disclosed in this manner may be combined in any combination. Unless otherwise expressly stated, each feature disclosed in this specification (including the accompanying claims, abstracts and drawings) may be replaced by an alternative feature providing the same, equivalent or similar purpose.

[0086] In addition, those skilled in the art will appreciate that, although some embodiments described herein include certain features included in other embodiments but not other features, the combination of features of different embodiments is meant to be within the scope of the embodiments of the present application and to form different embodiments. For example, in the claims below, any one of the claimed embodiments may be used in any combination.

[0087] The various component embodiments of the embodiments of the present application can be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. It should be understood by those skilled in the art that a microprocessor or digital signal processor (DSP) can be used in practice to implement some or all functions of some or all components according to the embodiments of the present application. The embodiments of the present application can also be implemented as a device or apparatus program (e.g., computer program and computer program product) for executing part or all of the methods described herein. Such a program implementing the embodiments of the present application can be stored on a computer-readable medium, or can have the form of one or more signals. Such a signal can be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

[0088] It should be noted that the above embodiments illustrate rather than limit the embodiments of the present application, and that those skilled in the art may design alternative embodiments without departing from the scope of the appended claims. In the claims, any reference symbol between brackets should not be constructed as a limitation to the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "one" or "an" preceding an element does not exclude the presence of multiple such elements. The embodiments of the present application may be implemented by means of hardware including several different elements and by means of appropriately programmed computers. In a unit claim that lists several devices, several of these devices may be embodied by the same hardware item. The use of the words first, second, and third, etc. does not indicate any order. These words may be interpreted as names.

Claims

1. A resource distribution method based on a causal model, characterized in that: include: Obtain historical behavior data of the target user, and predict the purchase amount range of the orders generated by the target user in the future based on the historical behavior data; Generate multiple resources within the purchase amount range and predict the use probability of each resource; Using a causal model, predict the residual of the evaluation index corresponding to each resource of the target user; According to the usage probability of each resource and the residual of the evaluation index corresponding to each resource, a resource distribution strategy is formulated for the target user.

2. The method according to claim 1, characterized in that The predicting, based on the historical behavior data, the purchase amount range of the orders generated by the target user in the future further includes: Preprocessing and feature extraction of the historical behavior data to determine corresponding correlation features; The associated features are input into an amount prediction model, and the amount prediction model predicts the purchase amount range of the orders generated by the target user in the future based on the associated features.

3. The method according to claim 2, characterized in that The associated features include: user features of the target user, features of commodities that have an interactive relationship with the target user, and cross-features between the target user and the commodities that have an interactive relationship with the target user.

4. The method according to claim 2, characterized in that: The step of generating a plurality of resources within the purchase amount range and predicting the usage probability of each resource further includes: Generating a plurality of resources in the purchase amount range according to the step length; The associated features are input into a probability prediction model, and the probability prediction model predicts the probability of use of each resource by the target user in the future based on the associated features.

5. The method according to any one of claims 1 to 4, characterized in that: The process of constructing the causal model includes: Obtaining the distribution data of each historical resource in the historical activity and the historical behavior data of each user in the historical activity; Divide each historical resource into multiple buckets, and build a causal model training data set based on the distribution data and the historical behavior data of each user; According to the causal model training data set, the residuals of the evaluation indicators corresponding to different users in each bucket are analyzed, and causal modeling is performed to obtain the causal model.

6. The method according to claim 5, characterized in that The use of the causal model to predict the residual of the evaluation index corresponding to each resource of the target user further includes: For each resource, according to the causal model, determine the bucket to which the resource belongs, similar users of the target user, and the residual of the evaluation index corresponding to the similar users in the bucket; The evaluation index residual corresponding to the target user under the resource is predicted based on the evaluation index residual corresponding to the similar user under the bucket and the similarity between the target user and the similar user.

7. A resource distribution device based on a causal model, characterized in that: include: A first prediction module is adapted to obtain historical behavior data of a target user and predict a purchase amount range of an order generated by the target user in the future based on the historical behavior data; A second prediction module, adapted to generate a plurality of resources within the purchase amount range and predict the use probability of each resource; A third prediction module is adapted to predict the residual of the evaluation index corresponding to each resource of the target user by using a causal model; The formulation module is adapted to formulate a resource distribution strategy for the target user according to the usage probability of each resource and the residual of the evaluation index corresponding to each resource.

8. A computing device comprising: A processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform operations corresponding to the resource distribution method based on a causal model as described in any one of claims 1-6.

9. A computer storage medium, wherein at least one executable instruction is stored in the storage medium, and the executable instruction enables a processor to execute operations corresponding to the resource distribution method based on a causal model as described in any one of claims 1 to 6.

10. A computer program product, comprising at least one executable instruction, wherein the executable instruction enables a processor to execute operations corresponding to the resource release method based on a causal model as described in any one of claims 1 to 6.

Images