User matching method and device, electronic equipment and computer readable medium

Through the incremental calculation model and Hungarian algorithm, intelligent matching between B-side account managers and customers is achieved, solving the problems of customer churn and marketing timing in the existing technology, and improving visit efficiency and customer retention.

CN119005624APending Publication Date: 2024-11-22BEIJING JINGDONG YUANSHENG TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411161855.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-22
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

The existing technology cannot conduct targeted intervention actions for B-end customers, resulting in customer churn and waste of marketing opportunities, lack of intelligent scheduling and optimization matching, resulting in low visit results and waste of manpower.

Method used

By building an incremental calculation model, combining user portrait data, business data, time data, logistics data, access traffic data and behavior data, the incremental value of issuance brought by each user to perform target operations on other users, and through Hungary The algorithm determines the optimal matching relationship and realizes intelligent matching between account managers and customers.

Benefits of technology

It improves visit efficiency and benefits, ensures customer retention and increases the order volume, finds out the best target actions through multi-dimensional analysis, and solves customer problems in a targeted manner.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119005624A_ABST
    Figure CN119005624A_ABST
Patent Text Reader

Abstract

The invention discloses a user matching method and device, electronic equipment and a computer readable medium, and relates to the technical field of big data processing and mining. The method comprises the following steps: inputting portrait data of each first user and comprehensive data of each second user into an increment calculation model corresponding to a dimension, thereby respectively obtaining an order issuing increment value brought by each first user to execute a target operation corresponding to the dimension on each second user; for each second user, adding the order issuing increment values brought by the fact that each first user executes the target operation corresponding to each dimension on the second user under each dimension, so as to respectively obtain the increment values brought by the fact that each first user executes the target operation corresponding to each dimension on the second user; and determining a one-to-one matching relationship between each first user and each second user according to the increment value brought by the execution of the target operation on each second user by each second user. According to the embodiment, the technical problem of lack of optimal matching between the customer and the customer manager can be solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of big data processing and mining technology, and in particular to a user matching method, device, electronic device and computer-readable medium. Background Art

[0002] At present, there is a relative lack of intelligent growth systems for the entire life cycle of B-side customers. The maintenance and operation of B-side customers is concentrated in the scope of sales process control, that is, the account manager subjectively judges the customer growth or loss node, and reaches the customer through offline visits, taps the potential of the customer or recovers the customer. Some intelligent growth systems mainly determine whether the customer is about to reach a certain marketing node by building a machine learning regression model or a time series model, so as to give the account manager corresponding action guidance and solve the single point problem of existing customers.

[0003] There are three main problems with the existing technology: First, it is impossible to carry out targeted intervention actions on customers in advance, resulting in customer loss and waste of marketing opportunities; second, there is a lack of intelligent scheduling of customer reach, the binding relationship between customers and account managers is relatively fixed, and the visiting time is relatively random. There is a lack of optimal matching between customers and account managers and dynamic optimal scheduling of visiting times, resulting in poor visiting effects and waste of manpower; third, there is a lack of multi-dimensional analysis of customers, resulting in unclear intervention actions and an inability to solve customer problems in a targeted manner. Summary of the invention

[0004] In view of this, embodiments of the present invention provide a user matching method, device, electronic device and computer-readable medium to solve the technical problem of lack of optimal matching between customers and account managers, resulting in poor visit results and waste of manpower.

[0005] To achieve the above object, according to one aspect of an embodiment of the present invention, a user matching method is provided, comprising:

[0006] For each dimension, the portrait data of each first user and the comprehensive data of each second user are input into the incremental calculation model corresponding to the dimension, so as to obtain the incremental value of orders caused by each first user performing the target operation corresponding to the dimension on each second user; wherein the comprehensive data includes at least one of the following: portrait data, business data, time data, logistics data, access flow data, and behavior data;

[0007] For each second user, the incremental values ​​of orders placed by each first user on the second user in each dimension are added together, thereby obtaining the incremental values ​​brought by each first user on the second user performing the target operations corresponding to each dimension;

[0008] According to the incremental values ​​brought by the target operations performed by the respective second users on the respective second users, a one-to-one matching relationship between the respective first users and the respective second users is determined to maximize the global total incremental value.

[0009] Optionally, the incremental calculation model corresponding to the dimension is trained using the following method:

[0010] A training sample set is constructed, and a regression model is used for supervised learning to construct an incremental calculation model corresponding to the dimension; wherein the training sample set includes portrait data of each first user, comprehensive data of each second user, feature data and order volume corresponding to the dimension, and comprehensive data and order volume of each second user.

[0011] Optionally, a training sample set is constructed, and a regression model is used for supervised learning, so as to construct an incremental calculation model corresponding to the dimension, including:

[0012] A training sample set is constructed, and a dual learner learning method is used to perform supervised learning on the regression model, and two sub-incremental models are obtained through training, thereby constructing an incremental calculation model corresponding to the dimension.

[0013] Optionally, a training sample set is constructed, and a T-learner method is used to perform supervised learning on the regression model, and two sub-incremental models are obtained through training, so as to construct an incremental calculation model corresponding to the dimension, including:

[0014] Constructing a first training sample set, and using a regression model for supervised learning, so as to train a first sub-incremental model; wherein the first training sample set includes the portrait data of each first user, the comprehensive data of each second user, the feature data corresponding to the dimension, and the order quantity;

[0015] Constructing a second training sample set, and using a regression model for supervised learning, thereby training a second sub-incremental model; wherein the second training sample set includes comprehensive data and order volume of each second user;

[0016] An incremental calculation model corresponding to the dimension is constructed according to the first sub-incremental model and the second sub-incremental model.

[0017] Optionally, for each dimension, the portrait data of each first user and the comprehensive data of each second user are input into the incremental calculation model corresponding to the dimension, so as to obtain the incremental value of orders caused by each first user performing the target operation corresponding to the dimension on each second user, including:

[0018] For each dimension, each first user, and each second user, input the portrait data of the first user and the comprehensive data of the second user into the first sub-incremental model, so as to output a first predicted order quantity resulting from the first user performing the target operation corresponding to the dimension on the second user;

[0019] For each dimension and each second user, the comprehensive data of the second user is input into the second sub-incremental model, so as to output a second predicted order quantity resulting from the first user not performing the target operation corresponding to the dimension on the second user;

[0020] Subtract the second predicted order quantity from the first predicted order quantity to obtain the incremental order value caused by the first user performing the target operation corresponding to the dimension on the second user.

[0021] Optionally, for each second user, the incremental values ​​of orders placed by each first user under each dimension when performing the target operation corresponding to each dimension on the second user are added together, so as to obtain the incremental values ​​brought by each first user performing the target operation corresponding to each dimension on the second user, respectively, including:

[0022] For each first user and each second user, the incremental value of orders placed by the first user performing the target operation corresponding to each dimension on the second user is added together to obtain the incremental value brought by the first user performing the target operation corresponding to each dimension on the second user.

[0023] Optionally, determining a one-to-one matching relationship between each first user and each second user according to an incremental value brought by each second user performing a target operation on each second user, so as to maximize a global total incremental value, includes:

[0024] According to the incremental value brought by each second user performing the target operation on each second user, the Hungarian algorithm is used to solve the one-to-one matching relationship between each first user and each second user, so as to maximize the global total incremental value.

[0025] In addition, according to another aspect of an embodiment of the present invention, a user matching device is provided, comprising:

[0026] A first calculation module is used for inputting the portrait data of each first user and the comprehensive data of each second user into the incremental calculation model corresponding to the dimension, so as to obtain the incremental value of orders caused by each first user performing the target operation corresponding to the dimension on each second user; wherein the comprehensive data includes at least one of the following: portrait data, business data, time data, logistics data, access flow data, and behavior data;

[0027] A second calculation module is used to add, for each second user, the incremental values ​​of orders placed by each first user under each dimension when performing the target operation corresponding to each dimension on the second user, thereby obtaining the incremental values ​​brought by each first user performing the target operation corresponding to each dimension on the second user;

[0028] The matching module is used to determine the one-to-one matching relationship between each first user and each second user according to the incremental value brought by each second user performing the target operation on each second user, so as to maximize the global total incremental value.

[0029] Optionally, the device further includes a training module for training an incremental calculation model corresponding to the dimension; wherein the incremental calculation model corresponding to the dimension is trained using the following method:

[0030] A training sample set is constructed, and a regression model is used for supervised learning to construct an incremental calculation model corresponding to the dimension; wherein the training sample set includes portrait data of each first user, comprehensive data of each second user, feature data and order volume corresponding to the dimension, and comprehensive data and order volume of each second user.

[0031] Optionally, the training module is further used to:

[0032] A training sample set is constructed, and a dual learner learning method is used to perform supervised learning on the regression model, and two sub-incremental models are obtained through training, thereby constructing an incremental calculation model corresponding to the dimension.

[0033] Optionally, the training module is further used to:

[0034] Constructing a first training sample set, and using a regression model for supervised learning, so as to train a first sub-incremental model; wherein the first training sample set includes the portrait data of each first user, the comprehensive data of each second user, the feature data corresponding to the dimension, and the order quantity;

[0035] Constructing a second training sample set, and using a regression model for supervised learning, thereby training a second sub-incremental model; wherein the second training sample set includes comprehensive data and order volume of each second user;

[0036] An incremental calculation model corresponding to the dimension is constructed according to the first sub-incremental model and the second sub-incremental model.

[0037] Optionally, the first calculation module is further used for:

[0038] For each dimension, each first user, and each second user, input the portrait data of the first user and the comprehensive data of the second user into the first sub-incremental model, so as to output a first predicted order quantity resulting from the first user performing the target operation corresponding to the dimension on the second user;

[0039] For each dimension and each second user, the comprehensive data of the second user is input into the second sub-incremental model, so as to output a second predicted order quantity resulting from the first user not performing the target operation corresponding to the dimension on the second user;

[0040] Subtract the second predicted order quantity from the first predicted order quantity to obtain the incremental order value caused by the first user performing the target operation corresponding to the dimension on the second user.

[0041] Optionally, the second calculation module is further used for:

[0042] For each first user and each second user, the incremental value of orders placed by the first user performing the target operation corresponding to each dimension on the second user is added together to obtain the incremental value brought by the first user performing the target operation corresponding to each dimension on the second user.

[0043] Optionally, the matching module is further used for:

[0044] According to the incremental value brought by each second user performing the target operation on each second user, the Hungarian algorithm is used to solve the one-to-one matching relationship between each first user and each second user, so as to maximize the global total incremental value.

[0045] According to another aspect of an embodiment of the present invention, there is also provided an electronic device, including:

[0046] one or more processors;

[0047] a storage device for storing one or more programs,

[0048] When the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any one of the above embodiments.

[0049] According to another aspect of an embodiment of the present invention, a computer-readable medium is provided, on which a computer program is stored. When the program is executed by a processor, the method described in any one of the above embodiments is implemented.

[0050] According to another aspect of an embodiment of the present invention, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the method described in any one of the above embodiments is implemented.

[0051] An embodiment of the above invention has the following advantages or beneficial effects: because for each second user, the incremental values ​​of orders brought by the target operations corresponding to each dimension performed by each first user on the second user are added together, so as to obtain the incremental values ​​brought by the target operations corresponding to each dimension performed by each first user on the second user, and then the one-to-one matching relationship between each first user and each second user is determined according to the incremental values ​​brought by the target operations performed by each second user on each second user, so as to maximize the global total incremental value, thus overcoming the technical problem of the lack of optimal matching between customers and account managers in the prior art. The embodiment of the present invention calculates the incremental values ​​brought by the target operations corresponding to each dimension performed by each first user on each second user through an incremental calculation model, and solves the matching relationship with the maximum global total incremental value, and intelligently matches the first user and the second user, so as to intelligently schedule the resources of the first user. The embodiment of the present invention not only makes the overall visit efficiency the highest and the visit revenue the largest, but also can perform multi-dimensional analysis on the second user, accurately find the best target action, so as to solve the problems of the second customer in a targeted manner, improve customer retention, and increase the number of customer orders.

[0052] The further effects of the above-mentioned non-conventional optional manner will be described below in conjunction with the specific implementation manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0054] Figure 1 is a flow chart of a user matching method according to an embodiment of the present invention;

[0055] Figure 2 is a schematic diagram of a bipartite graph of a first user and a second user according to an embodiment of the present invention;

[0056] Figure 3is a flow chart of a user matching method according to a reference embodiment of the present invention;

[0057] Figure 4 is a schematic diagram of a user matching device according to an embodiment of the present invention;

[0058] Figure 5 is an exemplary system architecture diagram to which embodiments of the present invention may be applied;

[0059] Figure 6 It is a schematic diagram of the structure of a computer system of a terminal device or a server suitable for implementing an embodiment of the present invention. DETAILED DESCRIPTION

[0060] The following is a description of exemplary embodiments of the present invention in conjunction with the accompanying drawings, including various details of the embodiments of the present invention to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0061] It should be noted that the collection, collection, updating, analysis, processing, use, transmission, storage and other aspects of user personal information involved in the technical solution of this disclosure are in compliance with the provisions of relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. Necessary measures are taken for user personal information to prevent illegal access to user personal information data and maintain the security of user personal information, network security and national security.

[0062] Figure 1 is a flow chart of a user matching method according to an embodiment of the present invention. As an embodiment of the present invention, Figure 1 As shown, the user matching method may include:

[0063] Step 101, for each dimension, input the portrait data of each first user and the comprehensive data of each second user into the incremental calculation model corresponding to the dimension, so as to obtain the incremental value of orders brought about by the target operation corresponding to the dimension performed by each first user on each second user; wherein the comprehensive data includes at least one of the following: portrait data, business data, time data, logistics data, access traffic data, and behavior data.

[0064] Before step 101, for each dimension, the incremental calculation model corresponding to each dimension is trained respectively, that is, each dimension has a corresponding incremental calculation model, through which the incremental calculation model can be used to calculate the incremental value of orders caused by each first user performing the target operation corresponding to the dimension on each second user under the dimension. Optionally, each dimension may include one or more of the equity dimension, the visit dimension, the loss dimension, the claim dimension, the price dimension, and the performance dimension. Optionally, the target operation may be to assign an equity budget (the target operation corresponding to the equity dimension), a visit action (the target operation corresponding to the visit dimension), solve the loss problem (the target operation corresponding to the loss dimension), solve the claim problem (the target operation corresponding to the claim dimension), solve the price problem (the target operation corresponding to the price dimension), and solve the performance problem (the target operation corresponding to the performance dimension).

[0065] Optionally, the incremental calculation model corresponding to the dimension is trained by the following method: construct a training sample set, and use a regression model for supervised learning to construct the incremental calculation model corresponding to the dimension; wherein the training sample set includes the portrait data of each first user, the comprehensive data of each second user, the feature data and the order volume corresponding to the dimension, and the comprehensive data and order volume of each second user. For each dimension, first construct a training sample set corresponding to the dimension, the training sample set includes the portrait data of each first user (such as years of work, job level, rights and interests budget that can be granted to the second user, the second user that can be visited, the loss problem that can be solved, the claim problem that can be solved, the price problem that can be solved, and the performance problem that can be solved), and the training sample set also includes the comprehensive data and order volume of each second user, wherein the comprehensive data of the second user includes portrait data, business data, time data, logistics data, access traffic data, and behavior data. Optionally, the second user's portrait data may include the current life cycle (such as the introduction period, the ramp-up period, the stable period, the decline period, the silent period, and the loss period, etc.), whether it is a POP merchant, the merchant size, the industry to which it belongs, the category to which the top consigned goods belong, the delivery cycle, the peak and off-peak season months of the industry, etc. Optionally, the second user's business data may include the number of orders issued in recent weeks (two weeks, three weeks, four weeks, or five weeks, etc.), the trend of the number of orders issued in recent weeks (two weeks, three weeks, four weeks, or five weeks, etc.), the number of orders issued in the same period in history, the average order volume, the month-on-month trend of the number of orders issued last month, the historical maximum number of orders issued, the historical average number of orders issued, the expiration time of the contract, the time of the most recent contract signing, the number of contracted products, the number of products used in recent weeks (two weeks, three weeks, four weeks, or five weeks, etc.), and the names of products used in recent weeks (two weeks, three weeks, four weeks, or five weeks, etc.). Optionally, the second user's time data may include the current month, the historical peak month for shipments, the historical valley month for shipments, etc. Optionally, the logistics data may include public logistics statistics, weather data, CPI, and GDP data. Optionally, the access traffic data of the second user may include access traffic data of other users in the store corresponding to the second user. Optionally, the behavior data of the second user may include the time when the second user performs the action of placing an order for shipment.

[0066] In an embodiment of the present invention, each dimension has its corresponding characteristic data. The characteristic data of the rights and interests dimension includes rights and interests-related data such as the rights and interests obtained by the second user and the amount of the rights and interests. The characteristic data of the visit dimension includes visit-related data such as the second user has been visited recently. The characteristic data of the loss dimension includes loss-related data such as the second user has recently suffered losses, loss data and loss amount. The characteristic data of the claim dimension includes claims-related data such as the second user has recently suffered claims, the number of claims, the claim amount, and the claim processing time limit. The characteristic data of the price dimension includes price-related data such as the second user's contract price, actual price, actual discount, and tiered discount. The characteristic data of the performance dimension includes performance-related data such as the second user's collection and delivery timeliness.

[0067] By performing 3sigma data denoising, continuous numerical standardization and category data embedding on the above multi-dimensional data, the processed features of different dimensions are spliced ​​and stored in the database for subsequent modeling and model prediction.

[0068] Optionally, a training sample set is constructed, and a regression model is used for supervised learning to construct an incremental computing model corresponding to the dimension, including: constructing a training sample set, and using a dual learner (T-learner) learning method to perform supervised learning on the regression model, training to obtain two sub-incremental models, and thus constructing an incremental computing model corresponding to the dimension. In an embodiment of the present invention, a T-learner is used to perform supervised learning on the regression model, and two sub-incremental models G and G' are trained respectively, and these two sub-incremental models constitute the incremental computing model corresponding to the dimension.

[0069] Optionally, step 101 may include: for each dimension, each first user and each second user, inputting the portrait data of the first user and the comprehensive data of the second user into the first sub-incremental model, thereby outputting a first predicted order quantity brought about by the first user performing the target operation corresponding to the dimension on the second user; for each dimension and each second user, inputting the comprehensive data of the second user into the second sub-incremental model, thereby outputting a second predicted order quantity brought about by the first user not performing the target operation corresponding to the dimension on the second user; subtracting the second predicted order quantity from the first predicted order quantity, thereby obtaining the incremental order value brought about by the first user performing the target operation corresponding to the dimension on the second user. After completing the incremental calculation model corresponding to each dimension, for each dimension, the portrait data of a first user and the comprehensive data of a second user are input into the first sub-incremental model, thereby outputting the first predicted order quantity brought about by the first user performing the target operation corresponding to the dimension on the second user. At the same time, the comprehensive data of the second user is input into the second sub-incremental model, thereby outputting the second predicted order quantity brought about by no first user performing the target operation corresponding to the dimension on the second user. Then, the second predicted order quantity is subtracted from the first predicted order quantity, thereby obtaining the incremental order value brought about by the first user performing the target operation corresponding to the dimension on the second user. ,Right now:

[0070]

[0071] Where X represents the feature vector, Indicates the order quantity. T=1 indicates that the first user performs the target operation corresponding to the dimension on the second user. T=0 indicates that no first user performs the target operation corresponding to the dimension on the second user.

[0072] like Figure 2 As shown, the bipartite graph can be used to obtain the gain value of each first user and each second user in each dimension. For dimension i, the gain value brought by the first user a performing the target operation corresponding to the dimension on the second user a can be calculated through step 101. , the gain value brought by the first user a performing the target operation corresponding to this dimension on the second user b , the gain value brought by the first user a performing the target operation corresponding to this dimension on the second user a , the gain value brought by the first user b performing the target operation corresponding to this dimension on the second user a , the gain value brought by the first user b performing the target operation corresponding to this dimension on the second user b , the gain value brought by the first user b performing the target operation corresponding to this dimension on the second user c , and so on, I will not go into details.

[0073] Step 102, for each second user, add up the incremental values ​​of orders brought about by each first user performing the target operation corresponding to each dimension on the second user in each dimension, so as to obtain the incremental values ​​brought about by each first user performing the target operation corresponding to each dimension on the second user.

[0074] Optionally, step 102 may include: for each first user and each second user, adding the incremental value of orders brought about by the first user performing the target operation corresponding to each dimension on the second user in each dimension, so as to obtain the incremental value brought about by the first user performing the target operation corresponding to each dimension on the second user.

[0075] like Figure 2 As shown, for the first user a and the second user a, the incremental value of the order placed by the first user a on the second user a in each dimension is added together to obtain the incremental value of the target operation corresponding to each dimension performed by the first user a on the second user a. :

[0076]

[0077] like Figure 2 As shown, for the first user a and the second user b, the incremental value of the order placed by the first user a on the second user b in each dimension is added together to obtain the incremental value of the target operation corresponding to each dimension performed by the first user a on the second user b. :

[0078]

[0079] like Figure 2 As shown, for the first user b and the second user a, the incremental value of the order placed by the first user b on the second user a in each dimension is added together, so as to obtain the incremental value of the target operation corresponding to each dimension performed by the first user b on the second user a. :

[0080]

[0081] Step 103: determining a one-to-one matching relationship between each first user and each second user according to the incremental value brought by each second user performing the target operation on each second user, so as to maximize the global total incremental value.

[0082] Optionally, step 103 may include: according to the incremental value brought by each second user performing the target operation on each second user, and solving by using the Hungarian algorithm, thereby solving the one-to-one matching relationship between each first user and each second user, so that the global total incremental value is maximized. After calculating the incremental value brought by each second user performing the target operation on each second user, solving by using the Hungarian algorithm, solving the matching relationship between the first user and the second user, so that the global total incremental value is maximized.

[0083] According to the various embodiments described above, it can be seen that the embodiment of the present invention adds the incremental values ​​of orders brought by the target operations corresponding to each dimension performed by each first user on the second user under each dimension, thereby obtaining the incremental values ​​brought by the target operations corresponding to each dimension performed by each first user on the second user, and then determines the one-to-one matching relationship between each first user and each second user according to the incremental values ​​brought by the target operations performed by each second user on each second user, so as to maximize the global total incremental value. The technical means solves the technical problem of the lack of optimal matching between customers and account managers in the prior art. The embodiment of the present invention calculates the incremental values ​​brought by the target operations corresponding to each dimension performed by each first user on each second user through an incremental calculation model, and solves the matching relationship with the largest global total incremental value, intelligently matches the first user and the second user, and thus intelligently schedules the resources of the first user. The embodiment of the present invention not only maximizes the overall visit efficiency and the visit revenue, but also can perform multi-dimensional analysis on the second user, accurately find the best target action, thereby solving the problems of the second customer in a targeted manner, improving customer retention, and increasing the number of customer orders.

[0084] Figure 3 is a flow chart of a user matching method according to a reference embodiment of the present invention. As another embodiment of the present invention, Figure 3 As shown, the user matching method may include:

[0085] Step 301: for each dimension, train an incremental calculation model corresponding to the dimension.

[0086] Optionally, the incremental calculation model corresponding to the dimension is trained by the following method: construct a first training sample set, and use a regression model for supervised learning, so as to train a first sub-incremental model; wherein the first training sample set includes the portrait data of each first user, the comprehensive data of each second user, the feature data corresponding to the dimension, and the order volume; construct a second training sample set, and use a regression model for supervised learning, so as to train a second sub-incremental model; wherein the second training sample set includes the comprehensive data and the order volume of each second user; the incremental calculation model corresponding to the dimension is constructed according to the first sub-incremental model and the second sub-incremental model. That is, the first training sample set includes the relevant data involved in the first user performing the target operation corresponding to the dimension on the second user, and the first training sample set includes the relevant data involved in the first user not performing the target operation corresponding to the dimension on the second user, and two sub-incremental models G and G' are trained respectively by these two training sample sets. That is, in the first training sample set, each sample includes the portrait data of the first user, the comprehensive data of the second user, the feature data corresponding to the dimension, and the order volume of the second user, and in the second training sample set, each sample includes the comprehensive data and the order volume of the second user.

[0087] In an embodiment of the present invention, T-learner is used to perform supervised learning on the regression model, and two sub-incremental models G and G' are trained respectively. The two sub-incremental models constitute the incremental calculation model corresponding to the dimension.

[0088] Step 302: For each dimension, each first user and each second user, the portrait data of the first user and the comprehensive data of the second user are input into the first sub-incremental model, thereby outputting the first predicted order quantity brought about by the first user performing the target operation corresponding to the dimension on the second user.

[0089] Step 303: For each dimension and each second user, the comprehensive data of the second user is input into the second sub-incremental model, thereby outputting a second predicted order quantity resulting from the first user not performing the target operation corresponding to the dimension on the second user.

[0090] Step 304: subtract the second predicted order quantity from the first predicted order quantity to obtain an incremental order value resulting from the first user performing the target operation corresponding to the dimension on the second user.

[0091] After completing the incremental calculation model corresponding to each dimension, for each dimension, the portrait data of a first user and the comprehensive data of a second user are input into the first sub-incremental model, thereby outputting the first predicted order quantity brought about by the first user performing the target operation corresponding to the dimension on the second user. At the same time, the comprehensive data of the second user is input into the second sub-incremental model, thereby outputting the second predicted order quantity brought about by no first user performing the target operation corresponding to the dimension on the second user. Then, the second predicted order quantity is subtracted from the first predicted order quantity, thereby obtaining the incremental order value brought about by the first user performing the target operation corresponding to the dimension on the second user. .

[0092] Step 305, for each first user and each second user, add up the incremental values ​​of orders brought about by the first user performing the target operations corresponding to each dimension on the second user in each dimension, so as to obtain the incremental values ​​brought about by the first user performing the target operations corresponding to each dimension on the second user.

[0093] Step 306: According to the incremental value brought by each second user performing the target operation on each second user, the Hungarian algorithm is used to solve the one-to-one matching relationship between each first user and each second user, so as to maximize the global total incremental value.

[0094] After respectively calculating the incremental values ​​brought by each second user performing the target operation on each second user, the Hungarian algorithm is used to solve the matching relationship between the first user and the second user so that the global total incremental value is maximized.

[0095] In addition, the specific implementation content of the user matching method in a reference embodiment of the present invention has been described in detail in the user matching method described above, so the repeated content will not be described here.

[0096] Figure 4 is a schematic diagram of a user matching device according to an embodiment of the present invention. Figure 4As shown, the user matching device 400 includes a first calculation module 401, a second calculation module 402 and a matching module 403; wherein the first calculation module 401 is used for, for each dimension, inputting the portrait data of each first user and the comprehensive data of each second user into the incremental calculation model corresponding to the dimension, so as to respectively obtain the incremental value of orders brought about by the target operation corresponding to the dimension performed by each first user on each second user; wherein the comprehensive data includes at least one of the following: portrait data, business data, time data, logistics data, access traffic data, and behavior data; the second calculation module 402 is used for, for each second user, adding the incremental value of orders brought about by the target operation corresponding to the dimension performed by each first user on the second user under each dimension, so as to respectively obtain the incremental value brought about by the target operation corresponding to the dimension performed by each first user on the second user; the matching module 403 is used for determining the one-to-one matching relationship between each first user and each second user according to the incremental value brought about by the target operation performed by each second user on each second user, so as to maximize the global total incremental value.

[0097] Optionally, the device further includes a training module for training an incremental calculation model corresponding to the dimension; wherein the incremental calculation model corresponding to the dimension is trained using the following method:

[0098] A training sample set is constructed, and a regression model is used for supervised learning to construct an incremental calculation model corresponding to the dimension; wherein the training sample set includes portrait data of each first user, comprehensive data of each second user, feature data and order volume corresponding to the dimension, and comprehensive data and order volume of each second user.

[0099] Optionally, the training module is further used to:

[0100] A training sample set is constructed, and a dual learner learning method is used to perform supervised learning on the regression model, and two sub-incremental models are obtained through training, thereby constructing an incremental calculation model corresponding to the dimension.

[0101] Optionally, the training module is further used to:

[0102] Constructing a first training sample set, and using a regression model for supervised learning, so as to train a first sub-incremental model; wherein the first training sample set includes the portrait data of each first user, the comprehensive data of each second user, the feature data corresponding to the dimension, and the order quantity;

[0103] Constructing a second training sample set, and using a regression model for supervised learning, thereby training a second sub-incremental model; wherein the second training sample set includes comprehensive data and order volume of each second user;

[0104] An incremental calculation model corresponding to the dimension is constructed according to the first sub-incremental model and the second sub-incremental model.

[0105] Optionally, the first calculation module 401 is further used for:

[0106] For each first user and each second user, input the portrait data of the first user and the comprehensive data of the second user into the first sub-incremental model, so as to output a first predicted order quantity resulting from the first user performing the target operation corresponding to the dimension on the second user;

[0107] For each dimension and each second user, the comprehensive data of the second user is input into the second sub-incremental model, so as to output a second predicted order quantity resulting from the first user not performing the target operation corresponding to the dimension on the second user;

[0108] Subtract the second predicted order quantity from the first predicted order quantity to obtain the incremental order value caused by the first user performing the target operation corresponding to the dimension on the second user.

[0109] Optionally, the second calculation module 402 is further used for:

[0110] For each dimension, each first user and each second user, the incremental value of orders placed as a result of the first user performing the target operation corresponding to each dimension on the second user is added together to obtain the incremental value of the target operation corresponding to each dimension performed by the first user on the second user.

[0111] Optionally, the matching module 403 is further configured to:

[0112] According to the incremental value brought by each second user performing the target operation on each second user, the Hungarian algorithm is used to solve the one-to-one matching relationship between each first user and each second user, so as to maximize the global total incremental value.

[0113] It should be noted that the specific implementation content of the user matching device of the present invention has been described in detail in the user matching method described above, so the repeated content will not be described again here.

[0114] Figure 5 An exemplary system architecture 500 is shown to which the user matching method or user matching device according to the embodiment of the present invention can be applied.

[0115] like Figure 5 As shown, system architecture 500 may include terminal devices 501, 502, 503, a network 504 and a server 505. Network 504 is used to provide a medium for communication links between terminal devices 501, 502, 503 and server 505. Network 504 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0116] Users can use terminal devices 501, 502, 503 to interact with server 505 through network 504 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 501, 502, 503, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social platform software, etc. (only for example).

[0117] The terminal devices 501 , 502 , and 503 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, and the like.

[0118] Server 505 may be a server that provides various services, such as a backend management server (only an example) that provides support for shopping websites browsed by users using terminal devices 501, 502, and 503. The backend management server may analyze and process the received data such as item information query requests, and feed back the processing results to the terminal device.

[0119] It should be noted that the user matching method provided in the embodiment of the present invention is generally executed by the server 505, and accordingly, the user matching device is generally set in the server 505. The user matching method provided in the embodiment of the present invention can also be executed by the terminal devices 501, 502, 503, and accordingly, the user matching device can be set in the terminal devices 501, 502, 503.

[0120] It should be understood that Figure 5 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0121] Reference below Figure 6 , which shows a schematic diagram of the structure of a computer system 600 of a terminal device suitable for implementing an embodiment of the present invention. Figure 6 The terminal device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0122] like Figure 6As shown, the computer system 600 includes a central processing unit (CPU) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage part 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the system 600 are also stored. The CPU 601, the ROM 602, and the RAM 603 are connected to each other through a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0123] The following components are connected to the I / O interface 605: an input section 606 including a keyboard, a mouse, etc.; an output section 607 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 608 including a hard disk, etc.; and a communication section 609 including a network interface card such as a LAN card, a modem, etc. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to the I / O interface 605 as needed. A removable medium 611, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 610 as needed, so that a computer program read therefrom is installed into the storage section 608 as needed.

[0124] In particular, according to the embodiments disclosed in the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present invention include a computer program, which includes a computer program carried on a computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 609, and / or installed from the removable medium 611. When the computer program is executed by the central processing unit (CPU) 601, the above-mentioned functions defined in the system of the present invention are executed.

[0125] It should be noted that the computer-readable medium shown in the present invention may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present invention, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0126] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, functions and operations of the system, method and computer program according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0127] The modules involved in the embodiments of the present invention may be implemented by software or hardware. The modules described may also be set in a processor, for example, it may be described as: a processor includes a first computing module, a second computing module and a matching module, wherein the names of these modules do not constitute a limitation on the modules themselves in some cases.

[0128] As another aspect, the present invention also provides a computer-readable medium, which may be included in the device described in the above embodiment; or it may exist independently and not be assembled into the device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by a device, the device implements the following method: for each dimension, the portrait data of each first user and the comprehensive data of each second user are input into the incremental calculation model corresponding to the dimension, so as to obtain the incremental value of the order issued by each first user when performing the target operation corresponding to the dimension on each second user; wherein the comprehensive data includes at least one of the following: portrait data, business data, time data, logistics data, access flow data, and behavior data; for each second user, the incremental value of the order issued by each first user when performing the target operation corresponding to each dimension on the second user under each dimension is added, so as to obtain the incremental value of the target operation corresponding to each dimension on the second user by each first user; according to the incremental value brought by the target operation performed by each second user on each second user, the one-to-one matching relationship between each first user and each second user is determined to maximize the global total incremental value.

[0129] As another aspect, an embodiment of the present invention further provides a computer program product, including a computer program, and when the computer program is executed by a processor, the method described in any of the above embodiments is implemented.

[0130] According to the technical solution of the embodiment of the present invention, for each second user, the incremental values ​​of orders brought by the target operations corresponding to each dimension performed by each first user on the second user are added together, thereby obtaining the incremental values ​​brought by the target operations corresponding to each dimension performed by each first user on the second user, and then determining the one-to-one matching relationship between each first user and each second user according to the incremental values ​​brought by the target operations performed by each second user on each second user, so as to maximize the global total incremental value. Therefore, the technical problem of the lack of optimal matching between customers and account managers in the prior art is overcome. The embodiment of the present invention calculates the incremental values ​​brought by the target operations corresponding to each dimension performed by each first user on each second user through an incremental calculation model, and solves the matching relationship with the maximum global total incremental value, and intelligently matches the first user and the second user, thereby intelligently scheduling the resources of the first user. The embodiment of the present invention not only makes the overall visit efficiency the highest and the visit revenue the largest, but also can perform multi-dimensional analysis on the second user, accurately find the best target action, thereby solving the problems of the second customer in a targeted manner, improving customer retention, and increasing the number of customer orders.

[0131] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions may occur depending on design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A user matching method, characterized in that: include: For each dimension, the portrait data of each first user and the comprehensive data of each second user are input into the incremental calculation model corresponding to the dimension, so as to obtain the incremental value of orders caused by each first user performing the target operation corresponding to the dimension on each second user; wherein the comprehensive data includes at least one of the following: portrait data, business data, time data, logistics data, access flow data, and behavior data; For each second user, the incremental values ​​of orders placed by each first user on the second user in each dimension are added together, thereby obtaining the incremental values ​​brought by each first user on the second user performing the target operations corresponding to each dimension; According to the incremental values ​​brought by the target operations performed by the respective second users on the respective second users, a one-to-one matching relationship between the respective first users and the respective second users is determined to maximize the global total incremental value.

2. The method according to claim 1, characterized in that The incremental calculation model corresponding to the dimension is trained using the following method: A training sample set is constructed, and a regression model is used for supervised learning to construct an incremental calculation model corresponding to the dimension; wherein the training sample set includes portrait data of each first user, comprehensive data of each second user, feature data and order volume corresponding to the dimension, and comprehensive data and order volume of each second user.

3. The method according to claim 2, characterized in that Construct a training sample set and use a regression model for supervised learning to construct an incremental calculation model corresponding to the dimension, including: A training sample set is constructed, and a dual learner learning method is used to perform supervised learning on the regression model, and two sub-incremental models are obtained through training, thereby constructing an incremental calculation model corresponding to the dimension.

4. The method according to claim 3, characterized in that A training sample set is constructed, and the regression model is supervisedly learned using the T-learner method to obtain two sub-incremental models, thereby constructing an incremental calculation model corresponding to the dimension, including: Constructing a first training sample set, and using a regression model for supervised learning, so as to train a first sub-incremental model; wherein the first training sample set includes the portrait data of each first user, the comprehensive data of each second user, the feature data corresponding to the dimension, and the order quantity; Constructing a second training sample set, and using a regression model for supervised learning, thereby training a second sub-incremental model; wherein the second training sample set includes comprehensive data and order volume of each second user; An incremental calculation model corresponding to the dimension is constructed according to the first sub-incremental model and the second sub-incremental model.

5. The method according to claim 4, characterized in that For each dimension, the portrait data of each first user and the comprehensive data of each second user are input into the incremental calculation model corresponding to the dimension, so as to obtain the incremental value of orders caused by each first user performing the target operation corresponding to the dimension on each second user, including: For each dimension, each first user, and each second user, input the portrait data of the first user and the comprehensive data of the second user into the first sub-incremental model, so as to output a first predicted order quantity resulting from the first user performing the target operation corresponding to the dimension on the second user; For each dimension and each second user, the comprehensive data of the second user is input into the second sub-incremental model, so as to output a second predicted order quantity resulting from the first user not performing the target operation corresponding to the dimension on the second user; Subtract the second predicted order quantity from the first predicted order quantity to obtain the incremental order value caused by the first user performing the target operation corresponding to the dimension on the second user.

6. The method according to claim 1, characterized in that For each second user, the incremental values ​​of orders placed by each first user under each dimension when performing the target operation corresponding to each dimension on the second user are added together, so as to obtain the incremental values ​​brought by each first user performing the target operation corresponding to each dimension on the second user, including: For each first user and each second user, the incremental value of orders placed by the first user performing the target operation corresponding to each dimension on the second user is added together to obtain the incremental value brought by the first user performing the target operation corresponding to each dimension on the second user.

7. The method according to claim 1, characterized in that Determining a one-to-one matching relationship between each first user and each second user according to the incremental value brought by each second user performing a target operation on each second user, so as to maximize the global total incremental value, includes: According to the incremental value brought by each second user performing the target operation on each second user, the Hungarian algorithm is used to solve the one-to-one matching relationship between each first user and each second user, so as to maximize the global total incremental value.

8. A user matching device, characterized in that: include: A first calculation module is used for inputting the portrait data of each first user and the comprehensive data of each second user into the incremental calculation model corresponding to the dimension, so as to obtain the incremental value of orders caused by each first user performing the target operation corresponding to the dimension on each second user; wherein the comprehensive data includes at least one of the following: portrait data, business data, time data, logistics data, access flow data, and behavior data; A second calculation module is used to add, for each second user, the incremental values ​​of orders placed by each first user under each dimension when performing the target operation corresponding to each dimension on the second user, thereby obtaining the incremental values ​​brought by each first user performing the target operation corresponding to each dimension on the second user; The matching module is used to determine the one-to-one matching relationship between each first user and each second user according to the incremental value brought by each second user performing the target operation on each second user, so as to maximize the global total incremental value.

9. An electronic device, characterized in that: include: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 7.

10. A computer readable medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

11. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.