A data processing method and device
By obtaining user characteristic information and using prediction models to determine the forwarding and conversion probability, and customizing the incentive information content, the problem of poor motivation information transmission effect in the existing technology is solved, and more efficient user motivation and promotion effects are achieved.
Patent Information
- Application Number
- CN202110524322.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-05-13
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2041-05-13
AI Technical Summary
The prior art lacks targetedness when sending incentive information to users and cannot effectively improve the forwarding probability and conversion probability.
By obtaining the characteristic information of the target user, based on the forwarding probability prediction model and the conversion probability prediction model, the user's forwarding probability and conversion probability are determined, and the user's coordinate information and level information are determined, and the content and method of the incentive information are customized based on the level information.
Targeted transmission of incentive information is achieved, which significantly improves the forwarding and conversion effect after the transmission of incentive information.
Smart Images

Figure CN113129080B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of computer technology, and in particular to a data processing method and device. Background Art
[0002] The user's forwarding probability and conversion probability are important factors affecting product promotion.
[0003] In the prior art, the forwarding probability is usually increased by sending incentive information to users, but the prior art only sends incentive information indiscriminately, and the effect produced after sending the incentive information is very poor. Summary of the invention
[0004] In view of this, an embodiment of the present invention provides a data processing method and device, which can send incentive information in a targeted manner and improve the effect produced after sending the incentive information.
[0005] In a first aspect, an embodiment of the present invention provides a data processing method, the method comprising:
[0006] Obtain characteristic information of target users;
[0007] According to the characteristic information of the target user, determining the forwarding probability of the target user based on a forwarding probability prediction model, and determining the conversion probability of the target user based on a conversion probability prediction model;
[0008] Determining the coordinate information of the target user according to the forwarding probability and the conversion probability;
[0009] Determining the level information of the target user according to the coordinate information;
[0010] Determining incentive information to be sent to the target user according to the level information;
[0011] The forwarding probability is used to characterize the probability that the target user makes a forwarding operation after receiving the incentive information, and the conversion probability is used to characterize the probability that the target user brings in new users after making the forwarding operation.
[0012] Further, the determining the coordinate information of the target user according to the forwarding probability and the conversion probability includes:
[0013] Determine the horizontal coordinate of the coordinate information according to the forwarding probability of the target user;
[0014] The vertical coordinate of the coordinate information is determined according to the conversion probability of the target user.
[0015] Further, determining the horizontal coordinate of the coordinate information according to the forwarding probability of the target user includes:
[0016] The difference between the forwarding probability of the target user and the forwarding probability threshold is used as the horizontal coordinate of the coordinate information;
[0017] Determining the ordinate of the coordinate information according to the conversion probability of the target user includes:
[0018] The difference between the conversion probability of the target user and the conversion probability threshold is used as the ordinate of the coordinate information.
[0019] Further, determining the level information of the target user according to the coordinate information includes:
[0020] In response to the abscissa of the coordinate information being positive and the ordinate of the coordinate information being positive, determining the level of the target user as the second level;
[0021] In response to the horizontal coordinate of the coordinate information being negative and the vertical coordinate of the coordinate information being positive, determining the level of the target user as the first level;
[0022] In response to the abscissa of the coordinate information being negative and the ordinate of the coordinate information being negative, determining the level of the target user as the fourth level; and
[0023] In response to the abscissa of the coordinate information being positive and the ordinate of the coordinate information being negative, the level of the target user is determined to be the third level.
[0024] Furthermore, the forwarding probability prediction model is trained in the following way:
[0025] Determining a first user set, the first user set including at least one user who has received over-stimulation information within a preset time period;
[0026] Determine a first training sample set according to the first user set;
[0027] The forwarding probability prediction model is trained according to the first training sample set.
[0028] Further, determining a first training sample set according to the first user set includes:
[0029] Determine a user in the first user set who has performed a forwarding operation after receiving the incentive information as a first positive sample user;
[0030] Determine a user in the first user set who has not performed a forwarding operation after receiving the incentive information as a first negative sample user;
[0031] The user features of the first positive sample user and the first negative sample user before receiving the incentive information are obtained to determine the first training sample set.
[0032] Furthermore, the conversion probability prediction model is trained in the following way:
[0033] Determine a first positive sample user set, wherein the first positive sample user set includes at least one of the first positive sample users;
[0034] Determine a second training sample set according to the first positive sample user set;
[0035] The conversion probability prediction model is trained according to the second training sample set.
[0036] Further, determining the second training sample set according to the first positive sample user set includes:
[0037] Determine a first positive sample user who brings a new user after the forwarding operation in the first positive sample user set as a second positive sample user;
[0038] Determine a first positive sample user in the first positive sample user set who does not bring in a new user after the forwarding operation as a second negative sample user;
[0039] User features of the second positive sample user and the second negative sample user before receiving the incentive information are obtained to determine a second training sample set.
[0040] In a second aspect, an embodiment of the present invention provides an electronic device, comprising a memory and a processor, wherein the memory is used to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method described in the first aspect.
[0041] In a third aspect, an embodiment of the present invention provides a computer-readable storage medium for storing computer program instructions, wherein the computer program instructions implement the method described in the first aspect when executed by a processor.
[0042] After acquiring the characteristic information of the target user, the method of the embodiment of the present invention determines the forwarding probability of the target user based on the forwarding probability prediction model according to the characteristic information of the target user, determines the conversion probability of the target user based on the conversion probability prediction model, determines the coordinate information of the target user according to the forwarding probability and the conversion probability, determines the level information of the target user according to the coordinate information, and determines the incentive information to be sent to the target user according to the level information. Through the method, the incentive information can be sent in a targeted manner to improve the effect of sending the incentive information. BRIEF DESCRIPTION OF THE DRAWINGS
[0043] The above and other objects, features and advantages of the present invention will become more apparent through the following description of the embodiments of the present invention with reference to the accompanying drawings, in which:
[0044] Figure 1 is a flow chart of a data processing method according to an embodiment of the present invention;
[0045] Figure 2 A schematic diagram of a process for determining target user coordinate information according to an embodiment of the present invention;
[0046] Figure 3 A schematic diagram of a process for determining target user level information according to an embodiment of the present invention;
[0047] Figure 4 A schematic diagram of a coordinate system according to an embodiment of the present invention;
[0048] Figure 5 A schematic diagram of a process of training a forwarding probability prediction model according to an embodiment of the present invention;
[0049] Figure 6 A schematic diagram of a process of training a conversion probability prediction model according to an embodiment of the present invention;
[0050] Figure 7 is a schematic diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION
[0051] The present invention is described below based on embodiments, but the present invention is not limited to these embodiments. In the detailed description of the present invention below, some specific details are described in detail. It is possible for a person skilled in the art to fully understand the present invention without the description of these details. In order to avoid confusing the essence of the present invention, known methods, processes, flows, components and circuits are not described in detail.
[0052] In addition, persons of ordinary skill in the art will appreciate that the drawings provided herein are for illustration purposes and are not necessarily drawn to scale.
[0053] Unless the context clearly requires otherwise, words such as “include”, “including” and the like in the specification should be interpreted as including rather than being exclusive or exhaustive; that is, as meaning “including but not limited to”.
[0054] In the description of the present invention, it should be understood that the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. In addition, in the description of the present invention, unless otherwise specified, "plurality" means two or more.
[0055] In the embodiments of the present invention, in order to protect the privacy of the user, the user information obtained is obtained with the permission of the user, and the user information is only applied to the methods described in the embodiments of the present invention.
[0056] In the process of promoting products, it is often necessary to encourage users to forward corresponding product promotion information to attract more new users.
[0057] In the prior art, in order to improve the promotion efficiency, incentive information is usually sent to users, and the incentive information includes rewards given to users and text content that encourages users to forward the information, so as to encourage users to forward the corresponding product promotion information. However, the effects of the incentive information received by different users are different. The prior art does not take into account the differences between users and cannot send corresponding incentive information to users in a targeted manner.
[0058] Figure 1 FIG. 1 is a flow chart of a data processing method according to an embodiment of the present invention. Figure 1 As shown, the data processing method of this embodiment includes the following steps.
[0059] S11: Obtain characteristic information of the target user.
[0060] The characteristic information includes the target user's age, historical forwarding times and types of purchased products.
[0061] Optionally, the characteristic information can be set according to actual needs. For example, the characteristic information can also include the number of times the target user purchases the product, the time of purchasing the product, the account registration time, the average forwarding frequency, etc. It should be understood that the characteristic information can also be combined in any way, such as only obtaining the user's age, the time of purchasing the product, and the account registration time.
[0062] S12: According to the characteristic information of the target user, determine the forwarding probability of the target user based on a forwarding probability prediction model, and determine the conversion probability of the target user based on a conversion probability prediction model.
[0063] Among them, the forwarding probability prediction model and the conversion probability prediction model are both pre-trained binary classification models. The forwarding probability is used to characterize the probability that the target user forwards the product information after receiving the incentive information, and the conversion probability is used to characterize the probability that the target user brings in new users after forwarding the product information.
[0064] Specifically, the characteristic information of the target user is respectively input into a pre-trained forwarding probability prediction model and a conversion probability prediction model, and the forwarding probability and the conversion probability of the target user are predicted by the prediction model.
[0065] Optionally, both the forwarding probability prediction model and the conversion probability prediction model may adopt an Xgboost model, and the forwarding probability and conversion probability of the target user may be predicted based on the Xgboost model.
[0066] Specifically, the Xgboost model consists of multiple CARTs (Classification and Regression Trees), each of which corresponds to different feature information, and the value corresponding to the leaf node of CART is an actual score.
[0067] In this embodiment, after receiving the training samples, the Xgboost model can find out the user features that affect the forwarding probability and the conversion probability from the training samples, and then establish the corresponding CART according to the found user features, and assign corresponding scores to the leaf nodes of each CART respectively, and obtain the forwarding probability prediction model and the conversion probability prediction model through training. The trained forwarding probability prediction model and the conversion probability prediction model can be classified and accumulated according to the input user features through each CART inside the model, and finally the scores obtained by each CART are added to obtain the forwarding probability and the conversion probability.
[0068] Optionally, Figure 5 The flowchart of the training forwarding probability prediction model according to the embodiment of the present invention can be as follows: Figure 5 The process shown trains the forwarding probability prediction model, which specifically includes the following steps.
[0069] S21: Determine a first user set.
[0070] The first user set includes at least one user who has received over-stimulation information within a preset time period.
[0071] Specifically, at least one user who has received incentive information within a preset time is determined as a first user set, wherein the number of users in the first user set can be set according to actual needs, wherein the incentive information includes rewards given to users, such as shopping coupons, and text content of promotional information that encourages users to forward the purchased products. The promotional information can be recommendation information of the product recommended by the user after using the purchased product, and the promotional information can also include a purchase link for the recommended product.
[0072] For example, if users A and B have received incentive information sent by product customer service within half a year, users A and B are determined as users in the first user set.
[0073] Optionally, the users in the first user set may be users who have received and read the motivational information within a preset time, and users who have received the motivational information but have not read it will not be determined in the first user set.
[0074] For example: users C and D have both received incentive information sent by product customer service within half a year. User C read the incentive information after receiving it, while user D failed to read the incentive information for some reason. At this time, only user C who received and read the incentive information is determined as a user in the first user set.
[0075] S22: Determine a first training sample set according to the first user set.
[0076] Optionally, the first training sample set may be determined through the following steps.
[0077] S221: Determine the users in the first user set who have performed a forwarding operation after receiving the incentive information as first positive sample users.
[0078] For example, after receiving the incentive information sent by the product customer service, user A in the first user set forwards the promotion information of the product he purchased to the social platform or shares it with his friends through the social platform, then user A is determined as the first positive sample user.
[0079] S222: Determine the users in the first user set who have not performed a forwarding operation after receiving the incentive information as first negative sample users.
[0080] For example, if user B in the first user set does not forward or share the promotion information after receiving the incentive information sent by the product customer service, user B is determined as the first negative sample user.
[0081] S223: Obtain user features of the first positive sample user and the first negative sample user before receiving incentive information to determine the first training sample set.
[0082] Specifically, after the users in the first user set are determined as first positive sample users and first negative sample users one by one, user features of the first positive sample users and the first negative sample users before receiving the incentive information are obtained to determine the first training sample set.
[0083] The user characteristics include at least the user's age, historical forwarding times and types of purchased products.
[0084] Optionally, the user characteristics may also be set to any characteristics that may affect the user's forwarding probability, such as: the number of times the product is purchased, the time when the product is purchased, the average forwarding frequency, and the last forwarding time.
[0085] S23: Training the forwarding probability prediction model according to the first training sample set.
[0086] Specifically, after obtaining the first training sample set, the first training sample set is input into the Xgboost model, and the Xgboost model is trained based on the first training sample set to find out the user features in the first training sample set that may affect the forwarding probability, and to establish multiple CARTs based on the found user features, and to assign corresponding scores to different leaf nodes of each CART. The trained Xgboost model can be used as a forwarding probability prediction model to predict the forwarding probability of the target user.
[0087] Optionally, Figure 6 The flowchart of the training conversion probability prediction model according to the embodiment of the present invention can be as follows: Figure 6 The process shown trains the conversion probability prediction model, which specifically includes the following steps.
[0088] S31: Determine a first positive sample user set.
[0089] The first positive sample user set includes at least one first positive sample user, and the first positive sample user is Figure 5 The first positive sample user in the illustrated process is specifically a user who has forwarded the information after receiving the incentive information within a preset time.
[0090] Optionally, the number of first positive sample users in the first positive sample user set can be set according to actual needs. For example, all first positive sample users can be determined as the first positive sample user set, or a part of the first positive sample users can be determined as the first positive sample user set.
[0091] S32: Determine a second training sample set based on the first positive sample user set.
[0092] Optionally, the second training sample set may be determined through the following steps.
[0093] S321: Determine a first positive sample user in the first positive sample user set who brings in a new user after a forwarding operation as a second positive sample user.
[0094] S322: Determine a first positive sample user in the first positive sample user set who does not bring in a new user after the forwarding operation as a second negative sample user.
[0095] For example: Users A and E are both first positive sample users. After user A forwards the promotional information of the product he purchased to the social platform, he brings in new users. At this time, user A is determined as the second positive sample user. However, after user E forwarded the product information, he failed to bring in new users. In this case, user E is determined as the second negative sample user.
[0096] S323: Obtain user features of the second positive sample user and the second negative sample user before receiving incentive information to determine a second training sample set.
[0097] There is a certain difference between the user features obtained in S323 and the user features obtained in S223. The difference is that the user features obtained in S323 are features that may affect the user conversion probability.
[0098] Optionally, the user features acquired in S323 and the user features acquired in S223 may partially overlap, or may be set separately.
[0099] For example, the user characteristics acquired in S323 may include the user's age, the type of product purchased, the number of historical conversions, the average conversion frequency, the account registration time, and the last conversion time, etc. Among them, the user's age and the type of product purchased overlap with the user characteristics acquired in S223, while the number of historical conversions, the average conversion frequency, the account registration time, and the last conversion time are set separately in S323.
[0100] S33: Training the conversion probability prediction model according to the second training sample set.
[0101] Specifically, after obtaining the second training sample set, the second training sample set is input into the Xgboost model, and the Xgboost model is trained based on the second training sample set to find out the user features in the second training sample set that may affect the conversion probability, and multiple CARTs are established based on the found user features, and corresponding scores are assigned to different leaf nodes of each CART. The trained Xgboost model can be used as a conversion probability prediction model to predict the conversion probability of the target user.
[0102] S13: Determine the coordinate information of the target user according to the forwarding probability and the conversion probability.
[0103] Optionally, Figure 2 The following is a flow chart of determining target user coordinate information according to an embodiment of the present invention. Figure 2 The process shown determines the coordinate information of the target user, and specifically includes the following steps.
[0104] S131: Determine the horizontal coordinate of the coordinate information according to the forwarding probability of the target user.
[0105] Specifically, the difference between the forwarding probability of the target user and the forwarding probability threshold is used as the horizontal coordinate of the coordinate information. The forwarding probability threshold can be set according to actual needs.
[0106] S132: Determine the vertical coordinate of the coordinate information according to the conversion probability of the target user.
[0107] Specifically, the difference between the conversion probability of the target user and the conversion probability threshold is used as the ordinate of the coordinate information. The conversion probability threshold can be set according to actual needs.
[0108] For example: after inputting the characteristic information of target user 1 into the forwarding probability prediction model and the conversion probability prediction model respectively, the forwarding probability and conversion probability of target user 1 are obtained to be 90% and 80% respectively, while the forwarding probability threshold and conversion probability threshold are set to be 80% and 70% respectively. At this time, (10, 10) can be used as the coordinate information of target user 1.
[0109] For another example: after the characteristic information of target user 2 is respectively input into the forwarding probability prediction model and the conversion probability prediction model, the forwarding probability and conversion probability of target user 2 are obtained to be 70% and 60% respectively, while the forwarding probability threshold and the conversion probability threshold are set to be 80% and 70% respectively. At this time, (-10, -10) can be used as the coordinate information of target user 2.
[0110] S14: Determine the level information of the target user according to the coordinate information.
[0111] Optionally, Figure 3 The flowchart of determining the target user level information according to an embodiment of the present invention can be as follows: Figure 3 The process shown determines the target user level information, and specifically includes the following steps.
[0112] S141: Determine whether the horizontal coordinate in the coordinate information is positive, if it is positive, execute S142, otherwise execute S145.
[0113] Optionally, if the horizontal coordinate in the coordinate information is 0, S142 is also executed.
[0114] S142: Determine whether the vertical coordinate in the coordinate information is positive, if it is positive, execute S143, otherwise execute S144.
[0115] Optionally, if the vertical coordinate in the coordinate information is 0, S143 is also executed.
[0116] S143: Determine the level of the target user as the second level.
[0117] Specifically, if both the horizontal and vertical coordinates of the target user's coordinate information are positive, it means that the forwarding probability of the target user is greater than or equal to the forwarding probability threshold, and the conversion probability is greater than or equal to the conversion probability threshold. That is, after sending the incentive information to the target user, the target user has a high probability of forwarding the promotional information of the product he purchased, and at the same time, there is a high possibility of bringing in new users after the target user forwards the promotional information. Among them, the larger the horizontal coordinate value in the coordinate information, the higher the probability of the target user forwarding, and the larger the vertical coordinate value, the higher the probability of the target user bringing in new users after forwarding the promotional information.
[0118] S144: Determine the level of the target user as the third level.
[0119] Specifically, if the horizontal coordinate of the target user's coordinate information is positive and the vertical coordinate is negative, it means that the forwarding probability of the target user is greater than or equal to the forwarding probability threshold, and the conversion probability is less than the conversion probability threshold. That is, after sending the incentive information to the target user, the target user has a high probability of forwarding the promotion information of the product he purchased, but there is only a small possibility of bringing in new users after the target user forwards the promotion information. Among them, the larger the horizontal coordinate value in the coordinate information, the higher the probability of the target user forwarding, and the smaller the vertical coordinate value, the lower the probability of the target user bringing in new users after forwarding the promotion information.
[0120] S145: Determine whether the vertical coordinate in the coordinate information is positive, if it is positive, execute S146, otherwise execute S147.
[0121] Optionally, if the vertical coordinate in the coordinate information is 0, S146 is also executed.
[0122] S146: Determine the level of the target user as the first level.
[0123] Specifically, if the horizontal coordinate of the target user's coordinate information is negative and the vertical coordinate is positive, it means that the forwarding probability of the target user is less than the forwarding probability threshold, and the conversion probability is greater than or equal to the conversion probability threshold. That is, after sending the incentive information to the target user, the target user is only very likely to forward the promotional information of the product he purchased, but there is a high probability of bringing in new users after the target user forwards the promotional information. Among them, the smaller the horizontal coordinate value in the coordinate information, the lower the probability of the target user forwarding, and the larger the vertical coordinate value, the higher the probability of the target user bringing in new users after forwarding the promotional information.
[0124] S147: Determine the level of the target user as the fourth level.
[0125] Specifically, if both the horizontal and vertical coordinates of the target user's coordinate information are negative, it means that the forwarding probability of the target user is less than the forwarding probability threshold, and the conversion probability is less than the conversion probability threshold. That is, after sending the incentive information to the target user, the target user is only very likely to forward the promotional information of the product he purchased, and at the same time, after the target user forwards the promotional information, it is only very likely to bring in new users. Among them, the smaller the horizontal coordinate value in the coordinate information, the lower the probability of the target user forwarding, and the smaller the vertical coordinate value, the lower the probability of the target user bringing in new users after forwarding the promotional information.
[0126] In an optional implementation, the target user's level information may be determined based on the forwarding probability threshold, the conversion probability threshold, and the target user's forwarding probability and conversion probability based on the four quadrants of the coordinate system. Specifically, the target user's level information may be determined by the following steps.
[0127] S41: Establishing a coordinate system according to the forwarding probability threshold and the conversion probability threshold.
[0128] Specifically, Figure 4 Schematic diagram of the coordinate system of an embodiment of the present invention, such as Figure 4 As shown, the coordinate system consists of a horizontal axis, a vertical axis and a center point O.
[0129] Among them, the horizontal axis is the forwarding probability, the vertical axis is the conversion probability, the horizontal axis of the center point O is the forwarding probability threshold, and the vertical axis is the conversion probability threshold. The specific values of the forwarding probability threshold and the conversion probability threshold can be set according to actual needs.
[0130] For example, when the forwarding probability threshold is set to 80% and the conversion probability threshold is set to 80%, the coordinate information of the center point O of the coordinate system is (80, 80).
[0131] S42: According to the forwarding probability and conversion probability of the target user, the target user is converted into a corresponding coordinate point in the coordinate system.
[0132] Specifically, the forwarding probability may be used as the horizontal coordinate and the conversion probability may be used as the vertical coordinate to convert the target user into a coordinate point in the coordinate system.
[0133] For example: the forwarding probability of target user 1 is 90%, and the conversion probability is 88%, then the corresponding coordinate point of target user 1 in the coordinate system is (90, 88); the forwarding probability of target user 2 is 86%, and the conversion probability is 76%, then the corresponding coordinate point of target user 2 in the coordinate system is (86, 76); the forwarding probability of target user 3 is 76%, and the conversion probability is 84%, then the corresponding coordinate point of target user 3 in the coordinate system is (76, 84); the forwarding probability of target user 4 is 70%, and the conversion probability is 74%, then the corresponding coordinate point of target user 4 in the coordinate system is (70, 74).
[0134] S43: Determine the level information of the target user according to the corresponding coordinate point of the target user in the coordinate system and based on the four quadrants of the coordinate system.
[0135] Specifically, Figure 4 As shown, the level of target user 1 in the first quadrant is determined as the second level, the level of target user 2 in the fourth quadrant is determined as the third level, the level of target user 3 in the second quadrant is determined as the first level, and the level of target user 4 in the third quadrant is determined as the fourth level.
[0136] Optionally, the level information of users on the boundary line between the first quadrant area and the second quadrant area is determined as the second level, the level information of users on the boundary line between the first quadrant area and the fourth quadrant area is determined as the second level, the level information of users on the boundary line between the second quadrant area and the third quadrant area is determined as the first level, and the level information of users on the boundary line between the third quadrant area and the fourth quadrant area is determined as the third level.
[0137] Compared to Figure 3 The method shown, through steps S41-S43, can process the levels of a large number of target users without calculating the coordinate information of the target users one by one and then making judgments based on the horizontal and vertical coordinates in the target information. Instead, a coordinate system is established to convert the target users into corresponding coordinate points in the coordinate system, and then the target users in different quadrants are obtained to determine the level information of the target users, thereby reducing the memory usage of the system.
[0138] S15: Determine the incentive information to be sent to the target user according to the level information.
[0139] Among them, the first-level users have a small probability of forwarding promotional information, but there is a high probability of bringing in new users after forwarding; the second-level users have a high probability of forwarding promotional information and bringing in new users; the third-level users have a low probability of forwarding promotional information and bringing in new users; the fourth-level users have a high probability of forwarding promotional information, but a low probability of bringing in new users after forwarding.
[0140] Specifically, the incentive information to be sent is determined according to the level information of the target user.
[0141] For example: if the level of target user 3 is the first level, then incentive information can be sent to target user 3 to encourage the user to forward the message. If the level of target user 1 is the second level, then incentive information may not be sent to target user 1 and the status quo may be maintained. If the levels of target user 2 and target user 4 are the third and fourth levels respectively, then corresponding preferential places may be given while incentive information is sent to target user 2 or target user 4. For example, when a new user purchases a product through the purchase link in the promotional information sent by target user 2 or target user 4, he or she may enjoy a certain discount, thereby increasing the probability of third-level and fourth-level users bringing in new users.
[0142] After acquiring the characteristic information of the target user, the method of the embodiment of the present invention determines the forwarding probability of the target user based on the forwarding probability prediction model according to the characteristic information of the target user, determines the conversion probability of the target user based on the conversion probability prediction model, determines the coordinate information of the target user according to the forwarding probability and the conversion probability, determines the level information of the target user according to the coordinate information, and determines the incentive information to be sent to the target user according to the level information. Through the method, the incentive information can be sent in a targeted manner to improve the effect of sending the incentive information.
[0143] Figure 7 Schematic diagram of an electronic device according to an embodiment of the present invention. Figure 7 As shown, the electronic device is a general data processing device, which includes a general computer hardware structure, which at least includes a processor 51 and a memory 52. The processor 51 and the memory 52 are connected via a bus 53. The memory 52 is suitable for storing instructions or programs executable by the processor 51. The processor 51 can be an independent microprocessor or a collection of one or more microprocessors. Thus, the processor 51 executes the instructions stored in the memory 52, thereby executing the method flow of the embodiment of the present invention as described above to realize the processing of data and the control of other devices. The bus 53 connects the above-mentioned multiple components together, and at the same time connects the above-mentioned components to the display controller 54 and the display device and the input / output (I / O) device 55. The input / output (I / O) device 55 can be a mouse, a keyboard, a modem, a network interface, a touch input device, a somatosensory input device, a printer, and other devices known in the art. Typically, the input / output device 55 is connected to the system via an input / output (I / O) controller 56.
[0144] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, devices (equipment) or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present application may adopt a computer program product implemented on one or more computer-readable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0145] The present application is described with reference to flowcharts of methods, apparatuses (devices) and computer program products according to embodiments of the present application. It should be understood that each process in the flowchart can be implemented by computer program instructions.
[0146] These computer program instructions may be stored in a computer-readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture including an instruction device that implements the process Figure 1 A function specified in a process or multiple processes.
[0147] These computer program instructions may also 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 produce the instructions for implementing the process Figure 1 A device that specifies functions in a process or multiple processes.
[0148] Another embodiment of the present invention relates to a non-volatile storage medium for storing a computer-readable program, wherein the computer-readable program is used for a computer to execute part or all of the above method embodiments.
[0149] That is, those skilled in the art can understand that all or part of the steps in the above-mentioned embodiment method can be completed by specifying relevant hardware through a program, and the program is stored in a storage medium, including several instructions for a device (which can be a single-chip microcomputer, chip, etc.) or a processor (processor) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.
[0150] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
Claims
1. A data processing method, characterized in that: The method comprises: Obtain characteristic information of target users; According to the characteristic information of the target user, the forwarding probability of the target user is determined based on a forwarding probability prediction model, and the conversion probability of the target user is determined based on a conversion probability prediction model, wherein the forwarding probability prediction model and the conversion probability prediction model are Xgboost models; Determining the coordinate information of the target user according to the forwarding probability and the conversion probability; Determining the level information of the target user according to the coordinate information; Determine, according to the level information, the incentive information to be sent to the target user and the preferential quota to be given to the target user; The forwarding probability is used to characterize the probability that the target user performs a forwarding operation after receiving the incentive information, and the conversion probability is used to characterize the probability that the target user brings in new users after performing the forwarding operation. The forwarding probability prediction model is obtained based on user feature training of users in the first user set before receiving the incentive information. The first user set includes at least one user who has received the incentive information within a preset time period. The conversion probability prediction model is obtained based on user feature training of the first positive sample user before receiving the incentive information. The first positive sample user is a user in the first user set who has performed a forwarding operation after receiving the incentive information. Wherein, determining the coordinate information of the target user according to the forwarding probability and the conversion probability includes: The difference between the forwarding probability of the target user and the forwarding probability threshold is used as the horizontal coordinate of the coordinate information; The difference between the conversion probability of the target user and the conversion probability threshold is used as the ordinate of the coordinate information; Wherein, determining the level information of the target user according to the coordinate information includes: In response to the abscissa of the coordinate information being positive and the ordinate of the coordinate information being positive, determining the level of the target user as the second level; In response to the horizontal coordinate of the coordinate information being negative and the vertical coordinate of the coordinate information being positive, determining the level of the target user as the first level; In response to the abscissa of the coordinate information being negative and the ordinate of the coordinate information being negative, determining the level of the target user as the fourth level; and In response to the horizontal coordinate of the coordinate information being positive and the vertical coordinate of the coordinate information being negative, determining the level of the target user as the third level; Determining the incentive information to be sent to the target user and the preferential quota to be given to the target user according to the level information includes: In response to the target user's level being the first level, sending incentive information to the target user; In response to the target user's level being the fourth level, a preferential quota is given to the target user.
2. The method according to claim 1, characterized in that The forwarding probability prediction model is trained in the following way: Determining a first user set; Determine a first training sample set according to the first user set; The forwarding probability prediction model is trained according to the first training sample set.
3. The method according to claim 2, characterized in that The determining, according to the first user set, a first training sample set comprises: Determine a user in the first user set who has performed a forwarding operation after receiving the incentive information as a first positive sample user; Determine a user in the first user set who has not performed a forwarding operation after receiving the incentive information as a first negative sample user; The user features of the first positive sample user and the first negative sample user before receiving the incentive information are obtained to determine the first training sample set.
4. The method according to claim 3, characterized in that The conversion probability prediction model is trained in the following way: Determine a first positive sample user set, wherein the first positive sample user set includes at least one of the first positive sample users; Determine a second training sample set according to the first positive sample user set; The conversion probability prediction model is trained according to the second training sample set.
5. The method according to claim 4, characterized in that The determining, according to the first positive sample user set, a second training sample set comprises: Determine a first positive sample user who brings a new user after the forwarding operation in the first positive sample user set as a second positive sample user; Determine a first positive sample user in the first positive sample user set who does not bring in a new user after the forwarding operation as a second negative sample user; User features of the second positive sample user and the second negative sample user before receiving the incentive information are obtained to determine a second training sample set.
6. An electronic device comprising a memory and a processor, characterized in that: The memory is used to store one or more computer program instructions, wherein the one or more computer program instructions are executed by the processor to implement the method according to any one of claims 1 to 5.
7. A computer-readable storage medium for storing computer program instructions, characterized in that: The computer program instructions implement the method according to any one of claims 1 to 5 when executed by a processor.
