CN110555747A - method and device for determining target user - Google Patents
method and device for determining target user Download PDFInfo
- Publication number
- CN110555747A CN110555747A CN201810558184.1A CN201810558184A CN110555747A CN 110555747 A CN110555747 A CN 110555747A CN 201810558184 A CN201810558184 A CN 201810558184A CN 110555747 A CN110555747 A CN 110555747A
- Authority
- CN
- China
- Prior art keywords
- purchase
- user
- target
- dimension
- commodity
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
Landscapes
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Engineering & Computer Science (AREA)
- General Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Marketing (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Theoretical Computer Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Data Mining & Analysis (AREA)
- Game Theory and Decision Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
the invention discloses a method and a device for determining a target user, and relates to the technical field of computers. One embodiment of the method comprises: training a purchase intention classification model by utilizing the value of a specific user in at least one preset primary selection dimension to obtain the weight value of each primary selection dimension; the specific user is a user who has performed purchase association behaviors aiming at the target commodity and then purchases the target commodity; determining the primary selection dimension with the weight value meeting the first judgment condition as a high-weight dimension, and acquiring the purchasing tendency score of a user who does not purchase any target commodity according to the high-weight dimension and the weight value thereof; and determining the users with the purchase tendency scores meeting the second judgment conditions as target users, and providing purchase guide objects of the target commodities for the target users. According to the embodiment, the target user can be accurately positioned through the purchase associated behavior of the user, and the purchase guide object is provided for the target user, so that the conversion of subsequent purchase behaviors is effectively improved.
Description
Technical Field
the present invention relates to the field of computer technologies, and in particular, to a method and an apparatus for determining a target user.
Background
In practical applications, it is often necessary to provide a target user with a purchase guide object (i.e., information or a real object for guiding a user to purchase a product), such as an advertisement, a trial product, etc., for the purpose of increasing the sales volume of the product. Since it costs some cost to push advertisement and send trial, it is necessary to precisely locate the target user with the appropriate demand to facilitate the subsequent purchase, thereby avoiding the loss of the service provider. In the prior art, the target user is generally located by analyzing characteristics of the user such as gender, age, consumption level, recent browsing behavior, recent shopping behavior and the like.
In the process of implementing the invention, the inventor finds that the prior art has at least the following problems:
1. The screening granularity of the characteristics is large, and the relevance of part of the characteristics and the shopping willingness of the user is weak, so that the range of the obtained target user is too large and the accuracy is low.
2. In the prior art, a positioning strategy can not be improved in a closed loop mode by utilizing feedback data of a user, and only limited adjustment can be carried out depending on manual experience.
Disclosure of Invention
In view of this, embodiments of the present invention provide a method and an apparatus for determining a target user, which can accurately locate the target user through a purchase correlation behavior of the user and provide a purchase guide object to the target user, so as to effectively improve a conversion from a purchase guide object to a purchase behavior.
To achieve the above object, according to one aspect of the present invention, there is provided a method of determining a target user.
the method for determining the target user in the embodiment of the invention comprises the following steps: training a pre-established purchase intention classification model by utilizing the values of a plurality of specific users in at least one preset primary selection dimension to obtain the weight value of each primary selection dimension; wherein the specific user is a user who has performed a purchase correlation action for the target commodity and thereafter purchases the target commodity; determining a primary selection dimension with a weight value meeting a preset first judgment condition as a high-weight dimension, and acquiring a purchasing tendency score of a user who does not purchase a target commodity according to the high-weight dimension and the weight value thereof; and determining the users with the purchase tendency scores meeting the preset second judgment conditions as target users, and providing purchase guide objects of the target commodities for the target users.
Optionally, the primary selection dimension is a virtual variable; and, the method further comprises: after determining the high weight dimension, for any user who has not purchased the destination item: and determining that the user has a high-weight dimension with a non-zero value, and taking a dimension option corresponding to the non-zero value as an portrait label of the user.
Optionally, the obtaining of the purchasing tendency score of the user who does not purchase the target product according to the high weight dimension and the weight value thereof specifically includes: and determining the sum of the weight values of the high-weight dimensions with non-zero values of the user as the purchase tendency score of the user.
Optionally, the purchase guidance object for providing the target product to the target product specifically includes: providing a target user with a purchase tendency score larger than a preset first threshold value with a purchase guide object of a target commodity; and providing the target users with the purchase tendency scores not larger than the first threshold value with the purchase guide objects and the promotion strategies of the target commodities.
Optionally, the method further comprises: after a purchase guide object of a target commodity is provided for a determined target user, training a current purchase intention classification model by using feedback data of the target user to adjust a weight value of the primary selection dimension; wherein the feedback data comprises: receiving the value of the target user who purchases the target commodity after purchasing the guide object in the primary selection dimension; determining the primary dimension of which the adjusted weight value meets a preset third judgment condition as a key dimension; for any user who does not purchase a target commodity, determining the sum of the weight values of the key dimensions with non-zero values of the user as the purchase tendency index of the user; and determining the users with the purchase tendency indexes meeting the preset fourth judgment conditions as target users, and providing purchase guide objects of the target commodities for the target users.
Optionally, the purchase correlation behavior comprises: searching for a commodity and not purchasing the commodity after a first period of time, accessing a commodity page and not purchasing the commodity after a second period of time, placing the commodity in a shopping cart and not purchasing the commodity after a third period of time or entering a commodity settlement page and not purchasing the commodity after a fourth period of time; the first discrimination condition includes: the weight value of the primary selection dimension is greater than a preset second threshold value or the weight value of the primary selection dimension is one of m maximum weight values; wherein m is a positive integer; the second determination condition includes: the purchasing tendency score of the user is between a preset third threshold value and a preset fourth threshold value; and the purchase intention classification model is a logistic regression model, and the purchase guide object is advertisement information or a trial product.
To achieve the above object, according to another aspect of the present invention, there is provided an apparatus for determining a target user.
The apparatus for determining a target user according to an embodiment of the present invention may include: the dimension weight obtaining unit is used for training a pre-established purchase intention classification model by utilizing the values of a plurality of specific users in at least one preset primary selection dimension to obtain the weight value of each primary selection dimension; wherein the specific user is a user who has performed a purchase correlation action for the target commodity and thereafter purchases the target commodity; the target user positioning unit is used for determining the primary selection dimension with the weight value meeting a preset first judgment condition as a high-weight dimension, and acquiring the purchasing tendency score of any user who does not purchase a target commodity according to the high-weight dimension and the weight value of the high-weight dimension; and determining the users with the purchase tendency scores meeting the preset second judgment conditions as target users, and providing purchase guide objects of the target commodities for the target users.
optionally, the primary selection dimension is a virtual variable; and, the apparatus may further comprise a portrait label determination unit for, after determining the high weight dimension, for a user of any non-purchased destination item: and determining that the user has a high-weight dimension with a non-zero value, and taking a dimension option corresponding to the non-zero value as an portrait label of the user.
optionally, the target user location unit may be further configured to: and determining the sum of the weight values of the high-weight dimensions with non-zero values of the user as the purchase tendency score of the user.
optionally, the target user location unit may be further configured to: providing a target user with a purchase tendency score larger than a preset first threshold value with a purchase guide object of a target commodity; and providing the target users with the purchase tendency scores not larger than the first threshold value with the purchase guide objects and the promotion strategies of the target commodities.
Optionally, the apparatus may further include a second iteration unit, configured to train a current purchasing intention classification model using the feedback data of the target user to adjust the weight value of the primary selection dimension after providing the purchase guide object of the target product to the determined target user; wherein the feedback data comprises: receiving the value of the target user who purchases the target commodity after purchasing the guide object in the primary selection dimension; determining the primary dimension of which the adjusted weight value meets a preset third judgment condition as a key dimension; for any user who does not purchase a target commodity, determining the sum of the weight values of the key dimensions with non-zero values of the user as the purchase tendency index of the user; and determining the users with the purchase tendency indexes meeting the preset fourth judgment conditions as target users, and providing purchase guide objects of the target commodities for the target users.
Optionally, the purchase correlation behavior may include: searching for a commodity and not purchasing the commodity after a first period of time, accessing a commodity page and not purchasing the commodity after a second period of time, placing the commodity in a shopping cart and not purchasing the commodity after a third period of time or entering a commodity settlement page and not purchasing the commodity after a fourth period of time; the first discrimination condition may include: the weight value of the primary selection dimension is greater than a preset second threshold value or the weight value of the primary selection dimension is one of m maximum weight values; wherein m is a positive integer; the second determination condition may include: the purchasing tendency score of the user is between a preset third threshold value and a preset fourth threshold value; the purchase intention classification model can be a logistic regression model, and the purchase guide object can be advertisement information or a trial product.
to achieve the above object, according to still another aspect of the present invention, there is provided an electronic apparatus.
An electronic device of the present invention includes: one or more processors; a storage device for storing one or more programs which, when executed by the one or more processors, cause the one or more processors to implement the method of determining a target user provided by the present invention.
To achieve the above object, according to still another aspect of the present invention, there is provided a computer-readable storage medium.
A computer-readable storage medium of the present invention has stored thereon a computer program which, when executed by a processor, implements the method of determining a target user provided by the present invention.
according to the technical scheme of the invention, one embodiment of the invention has the following advantages or beneficial effects:
Firstly, a specific user who executes the purchasing associated behavior is determined, the value of the specific user in the primary selection dimension is input into a purchasing intention classification model to obtain the weight value of the primary selection dimension, then the high-weight dimension with strong correlation with the purchasing intention of the user can be obtained by utilizing the weight value, the purchasing tendency score of the user can be calculated, and finally the target user is positioned. Through the steps, the target user can be accurately positioned based on the purchase association behavior which directly reflects the user intention, and the defects of low positioning accuracy and large screening granularity in the prior art are overcome.
secondly, after the purchase guide object is provided for the user for the first time, the purchase intention classification model can be further trained by utilizing feedback data, so that the weight value of each primary selection dimension is adjusted, a key dimension which is more accurate compared with a high-weight dimension is obtained, the target user can be more accurately positioned through the key dimension, and the purchase guide object is provided for the second time (the second time is different from the first-time targeted user). And then, the weight value of the primary selection dimension can be adjusted again by utilizing the feedback data for the second time so as to provide a purchase guide object for the third time.
Thirdly, the present invention can set various preset conditions according to the application scenario to flexibly adjust the high-weight dimension, the key dimension and the range of the target user, and can selectively execute various actions (such as providing a purchase guide object, a promotion strategy, etc.) according to the purchase tendency score or the purchase tendency index of the target user, thereby generating good subsequent purchase conversion.
Fourthly, the invention sets the primary dimension as a virtual variable (Dummy Variables) to reduce the operation amount and enhance the system robustness.
further effects of the above-mentioned non-conventional alternatives will be described below in connection with the embodiments.
drawings
The drawings are included to provide a better understanding of the invention and are not to be construed as unduly limiting the invention. Wherein:
FIG. 1 is a schematic diagram of the main steps of a method for determining a target user according to an embodiment of the present invention;
FIG. 2 is a diagram illustrating an implementation of a method for determining a target user according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of the components of an apparatus for identifying a target user in an embodiment in accordance with the invention;
FIG. 4 is an exemplary system architecture diagram in which embodiments of the present invention may be employed;
Fig. 5 is a schematic structural diagram of an electronic device for implementing the method for determining the target user in the embodiment of the present invention.
Detailed Description
exemplary embodiments of the present invention are described below with reference to the accompanying drawings, in which various details of embodiments of the invention are included to assist understanding, and which are to be considered as merely exemplary. Accordingly, those of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the invention. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.
It should be noted that the embodiments of the present invention and the technical features of the embodiments may be combined with each other without conflict.
Fig. 1 is a schematic diagram of the main steps of a method for determining a target user according to an embodiment of the present invention.
as shown in fig. 1, the method for determining a target user according to the embodiment of the present invention may be specifically executed according to the following steps:
Step S101: and training a pre-established purchase intention classification model by utilizing the values of a plurality of specific users in at least one preset primary selection dimension to obtain the weight value of each primary selection dimension.
In this step, the specific user refers to a user who has performed the purchase association action for the target commodity once and purchased the target commodity after a certain period of time. The target product, which is an object product, may be any specific product as an application object of the method of the present invention. Illustratively, the purchase correlation behavior may be any of the following: the method includes the steps of searching for a commodity and not purchasing the commodity after a first time period (may be referred to as a search non-purchase behavior for short), accessing a commodity page and not purchasing the commodity after a second time period (may be referred to as an access non-purchase behavior for short), placing the commodity in a shopping cart and not purchasing the commodity after a third time period (may be referred to as a shopping cart non-purchase behavior for short), and entering a commodity settlement page and not purchasing the commodity after a fourth time period (may be referred to as an entry settlement page non-purchase behavior for short), wherein the above time periods can be flexibly set according to actual scenes.
In a specific application, the actual scenario of searching for unpurchased behavior may be: the user listens to a certain commodity but does not know specific information thereof, and searches for it for curiosity. The actual scenario of accessing unpurchased behavior may be: the user would like to fully understand the detailed information of a certain commodity, but abandon the purchase because the function of the commodity cannot completely match the requirement of the commodity or because the price of the commodity is higher. The actual scenario of an unpurchased action of placing a shopping cart may be: the user is mindful of a commodity and has accepted its price, but abandons the commodity because a more suitable alternative appears. The actual scenario of an unpurchased action into a checkout page may be: the user is mindful of a certain item and accepts its price while there are no alternatives, but the item has to be temporarily abandoned since the total price of several items in an order is over the budget. It can be seen that the above four purchase association behaviors can directly reflect the purchase intentions of users with different degrees, and the specific users who have executed the purchase association behaviors and finally make purchases at the later stage are screened out and the characteristics of each specific user are analyzed, so that the specific attributes of the users who are easy to realize purchase conversion (i.e. target users) can be outlined.
generally, the primary selection dimension may be preset according to an application environment. For example, gender, marital status, etc. may be set. In particular, in the embodiment of the present invention, in order to reduce the amount of operations and enhance the system robustness, each of the primary dimensions may be set as a virtual variable (i.e., a dummy variable) whose value can only be 0 or 1. In a specific application, a user characteristic with a plurality of values can be converted into a plurality of virtual variables. For example: the Age characteristics can be converted into the following five primary dimensions Age _1, Age _2, Age _3, Age _4, Age _5, etc.:
wherein, the dimension option represents a plurality of options of a certain primary selection dimension. For example: the dimension options for the "gender" dimension are "male" and "female". A certain value of the user in the initial selection dimension corresponds to a dimension option, and the classification contribution of the dimension option can be reflected.
Although the virtual variables are used as examples in the above, the initial dimensions of the present invention are not limited to the virtual variables. In fact, the initial selection vector may be a variable with any value, and the present invention is not limited in this respect.
in one embodiment, the pre-established purchase intention classification model indicates that the mathematical structure of the model is pre-established, but the model parameters need to be determined through a subsequent training process. The purchase intention classification model is used for judging the purchase probability according to the value of the user in the primary selection dimension, and can be a classification model such as naive Bayes, a decision tree, logistic regression and the like. In this step, a specific user who has performed a search unpurchased behavior, accessed an unpurchased behavior, placed in a shopping cart unpurchased behavior, or entered a settlement page unpurchased behavior may be selected first, and then a value of a certain class of specific users in at least one primary selection dimension is input into the purchase intention classification model for training to obtain a weight value of each primary selection dimension, where the weight value may reflect a classification weight of the primary selection dimension in the purchase intention classification model. Obviously, 4 kinds of purchase intention classification models with different parameters are available for the above 4 kinds of specific users.
The following description will be made by using a logistic regression model. In the logistic regression model, the logistic function is:
Wherein, X is an independent variable, namely a vector formed by values of the initially selected dimension of a specific user; y is a dependent variable and can be 1 or 0, wherein 1 represents purchase after the purchase association behavior is executed, and 0 represents purchase after the purchase association behavior is executed; theta is a vector formed by weighted values corresponding to the primary dimension to be solved, and T represents transposition.
Then, an objective function can be established by applying a maximum likelihood estimation method, and a weight value of each initially selected dimension is determined by utilizing a Newton-Raphson method. The following table shows a plurality of primary dimensions obtained by converting the age characteristics and the weight values thereof determined according to the logistic regression model.
The following table shows values, dimension options and primary dimension weight values of a plurality of primary dimensions obtained by converting the characteristic of 'purchase amount of a user in 3 months'.
step S202: determining a primary selection dimension with a weight value meeting a preset first judgment condition as a high-weight dimension, and acquiring a purchasing tendency score of a user who does not purchase a target commodity according to the high-weight dimension and the weight value thereof; and determining the users with the purchase tendency scores meeting the preset second judgment conditions as target users, and providing purchase guide objects of the target commodities for the target users.
In this step, the first determination condition and the second determination condition may be set according to the application environment. For example, the first discrimination condition may be set to: the weight value of the primary selection dimension is greater than a preset second threshold or one of m (which is a positive integer) weight values of which the weight value of the primary selection dimension is the maximum, that is, the high-weight dimension can be determined according to the following two ways:
the first mode is as follows: taking the primary selection dimension with the weight value larger than a second threshold value as a high weight dimension;
The second mode is as follows: and (4) sorting the weighted values of all the primary selection dimensions in a descending order, and taking the previous m primary selection dimensions as high-weighted dimensions.
And then, acquiring the purchase tendency score of any user who does not purchase the target commodity according to the high weight dimension and the weight value thereof, wherein the purchase tendency score can be used for representing the purchase conversion probability of the user who does not purchase the target commodity. Specifically, the product of the value of the user in each high-weight dimension and the weight value of the high-weight dimension can be calculated, and the sum of all the products is used as the purchase tendency score of the user. If the primary selection dimension is a virtual variable, the purchase tendency score of the user is equal to the sum of the weight values of the high-weight dimensions with non-zero values of the user.
As a preferred scheme, after the high-weight dimension is obtained, for any user who does not purchase a target product, the high-weight dimension with a non-zero value is determined, and a dimension option corresponding to the non-zero value is used as an image label of the user, so that the user with a strong purchase intention is visually marked, and guidance is provided for making a subsequent sales strategy. For example: if the primary dimensions "Age _ 3" and "Sum _ 4" in the previous table are high-weight dimensions, and the values of the user who has not purchased the target product in the two dimensions are both 1 (non-zero values), the corresponding dimension options "35 < Age ≦ 40" and "purchase amount in 3 months >6 ten thousand" may be used as the labels of the user figures.
in one embodiment, after obtaining the purchasing tendency score of the user, the user whose purchasing tendency score meets the second criterion may be determined as the target user. For example: when the resources are abundant, users with purchase tendency scores larger than a preset third threshold value can be taken as target users; when the resources are strained, users with a purchasing propensity score between the third threshold and the fourth threshold (i.e., purchasing propensity scores greater than the third threshold and less than the fourth threshold) may be targeted. This is because: among users with purchase tendency scores larger than the third threshold, users not smaller than the fourth threshold have stronger purchase intentions, and the subsequent purchase conversion cannot be greatly influenced even if the users are not taken as target users (namely, subsequent actions are not performed on the users), so that the users can be excluded from the target users when resources are in shortage; and the users with the purchase tendency scores between the third threshold and the fourth threshold, which have slightly small purchase willingness, can be used as the subjects of the target users and perform subsequent actions specifically to generate purchase conversion.
It is understood that the second determination condition may be set to take a certain percentage of users as target users. For example: when the resources are abundant, the users with the largest purchasing tendency scores of 30% can be used as target users; when the resources are short, the user with the highest tendency score of 5% (the proportion of the user as a whole) may be excluded from the above users, and the remaining users may be the target users.
then, a purchase guide object of the target item (i.e., advertisement information of the target item or a trial) may be provided to the target user to prompt the target user to purchase the target item. It can be understood that the purchasing associated behaviors of the user are mined in the steps, so that the high-weight dimension and the weight value thereof are determined, and the purchasing tendency score of the user is calculated according to the high-weight dimension and the weight value to realize the accurate evaluation of the purchasing desire of the user.
in practical applications, different actions may be performed for target users with different purchasing propensity scores. For example: for target users with purchase tendency scores larger than a preset first threshold, only providing a purchase guide object of a target commodity (because the purchase intention of the target users is strong, only providing the purchase guide object can promote the purchase of the target users); for target users with the purchase tendency scores less than or equal to the first threshold, the purchase guidance object and the promotion strategy of the target products are provided (because the purchase intention of such target users is weaker, the promotion strategies of discounting prices, reaching a certain amount of money, giving off benefits, returning vouchers and the like are further provided to accelerate the purchase conversion of the target products).
Through the steps S101 and S102, the target user can be accurately positioned based on the purchase association behavior directly reflecting the user intention, and the purchase guide object is provided to promote the final purchase, so that the problems of low positioning accuracy and large screening granularity in the prior art are solved.
preferably, in an embodiment, after the purchase guidance object is provided to the target user, feedback data of the target user is obtained, and the weight value of the primary dimension is adjusted by using the feedback data. The feedback data refers to a value of a primary selection dimension of a target user who purchases a target commodity after receiving the purchase guide object.
Specifically, the current purchase intention classification model is further trained by using feedback data to adjust the weight value of the primary dimension, and then the primary dimension of which the adjusted weight value meets a preset third judgment condition is determined as the key dimension. It will be appreciated that the third discrimination condition may be set in a manner similar to the first discrimination condition (only the specific threshold needs to be changed). Since the key dimension is obtained by optimizing the feedback data on the basis of the high-weight dimension, the reliability of the key dimension is higher than that of the high-weight dimension. The key dimensions and their current weight values may then be used to calculate a purchase propensity index (which may function similarly to the purchase propensity score, except that the dimensions used in the calculation may be different) for any user who has not purchased the target item. During specific calculation, if the key dimension is a virtual variable, the sum of the weight values of the key dimensions with non-zero values of the user can be determined as the purchase tendency index. It can be understood that after the key dimension is obtained, the key dimension of the user with the non-zero value can be determined for any user who does not purchase the target commodity, and the dimension option corresponding to the non-zero value is used as the portrait label of the user to guide the formulation of the subsequent sales strategy.
After obtaining the purchase tendency index of the user, the user whose purchase tendency index meets the preset fourth determination condition may be determined as the target user, and the purchase guide object of the target product may be provided thereto. Wherein the fourth discrimination condition acts similarly to the second discrimination condition, and the difference therebetween lies only in the setting of the specific threshold. It is understood that, since the target user at this time is determined based on the feedback data representing the actual purchase situation, the accuracy thereof is higher than that of the target user determined in steps S101 and S102. Through the arrangement, further accurate positioning of the target user and closed-loop optimization of the positioning strategy can be realized.
It will be appreciated that after the second providing of the purchase leader, the corresponding feedback data may continue to be obtained to similarly identify the more accurate target user and provide the purchase leader again, after which the feedback data is obtained again and the purchase leader … … is provided again to repeat the above steps to achieve an iterative optimization of the positioning strategy until the predetermined termination condition is met. Therefore, the invention gradually and clearly outlines the specific attributes of the target user through the multiple iterative processes, thereby effectively promoting the subsequent purchase of the commodity and solving the problem that the positioning strategy can not be improved by utilizing the feedback data closed loop in the prior art.
It should be noted that, in the actual user location process, the purchase intention classification model may be any one of the following models: the model based on the search unpurchased behavior (i.e., the model is trained by the relevant data of the specific user who performed the search unpurchased behavior), the model based on the access unpurchased behavior (i.e., the model is trained by the relevant data of the specific user who performed the access unpurchased behavior), the model based on the built-in shopping cart unpurchased behavior (i.e., the model is trained by the relevant data of the specific user who performed the built-in shopping cart unpurchased behavior), the model based on the built-in settlement page unpurchased behavior (i.e., the model is trained by the relevant data of the specific user who performed the built-in settlement page unpurchased behavior), and any combination of 4 models, for: and respectively setting discrimination conditions for the model based on the search unpurchased behavior and the model based on the access unpurchased behavior to determine a target user, and respectively acquiring corresponding feedback data optimization models after providing a purchase guide object for the target user. The following steps are repeated: discrimination conditions can be respectively set for the 4 models to jointly determine a target user, and feedback data of each model is respectively obtained after the target user is provided with a purchased guide object to optimize the corresponding model.
Fig. 2 is a schematic diagram of a specific implementation of the method for identifying a target user according to the embodiment of the present invention, wherein the steps of identifying a target user and providing a trial product using the 4 models are shown. Specifically, the purchasing behavior data and the order data of the user are firstly cleaned to obtain the 4 specific users and are modeled by the related data of the specific users, and then the high-weight dimension is obtained from the purchasing intention classification model to further determine the purchasing tendency score of the user, so that the target user can be positioned and the trial products can be issued to the target user, and the first issuing of the trial products is completed. And then, when the trial products have surplus, feedback data of the target user is acquired and fed back to the purchase intention classification model after cleaning so as to adjust the weight value of the primary selection dimension, and therefore the more accurate target user is determined. And repeating the iteration process to realize continuous closed-loop optimization of the target user positioning strategy.
According to the technical scheme of the embodiment of the invention, the high-weight dimension and the key dimension can be extracted according to the purchase association behavior of the user, so that the accurate positioning of the target user is realized, the user positioning strategy can be adjusted in a closed loop according to the feedback data, and the effective promotion of the commodity sales volume is finally realized.
Fig. 3 is a schematic diagram of a part of an apparatus for determining a target user according to an embodiment of the present invention.
as shown in fig. 3, an apparatus 300 for determining a target user according to an embodiment of the present invention may include a dimension weight obtaining unit 301 and a target user locating unit 302. Wherein:
the dimension weight obtaining unit 301 may be configured to train a pre-established purchase intention classification model by using values of a plurality of specific users in at least one preset primary selection dimension, and obtain a weight value of each primary selection dimension; wherein the specific user is a user who has performed a purchase correlation action for the target commodity and thereafter purchases the target commodity;
the target user positioning unit 302 may be configured to determine a primary selection dimension with a weight value meeting a preset first determination condition as a high-weight dimension, and obtain a purchasing tendency score of a user who does not purchase a target product according to the high-weight dimension and the weight value thereof; and determining the users with the purchase tendency scores meeting the preset second judgment conditions as target users, and providing purchase guide objects of the target commodities for the target users.
In the embodiment of the invention, the primary selection dimension is a virtual variable; and, the apparatus 300 may further comprise a portrait label determination unit for, after determining the high weight dimension, for a user of any non-purchased destination item: and determining that the user has a high-weight dimension with a non-zero value, and taking a dimension option corresponding to the non-zero value as an portrait label of the user.
As a preferred solution, the target user location unit 302 may be further configured to: and determining the sum of the weight values of the high-weight dimensions with non-zero values of the user as the purchase tendency score of the user.
Preferably, in one embodiment, the target user location unit 302 is further configured to: providing a target user with a purchase tendency score larger than a preset first threshold value with a purchase guide object of a target commodity; and providing the target users with the purchase tendency scores not larger than the first threshold value with the purchase guide objects and the promotion strategies of the target commodities.
In practical applications, the apparatus 300 may further include a second iteration unit, configured to train a current purchasing intention classification model using the feedback data of the target user to adjust the weight value of the primary selection dimension after providing the purchase guide object of the target product to the determined target user; wherein the feedback data comprises: receiving the value of the target user who purchases the target commodity after purchasing the guide object in the primary selection dimension; determining the primary dimension of which the adjusted weight value meets a preset third judgment condition as a key dimension; for any user who does not purchase a target commodity, determining the sum of the weight values of the key dimensions with non-zero values of the user as the purchase tendency index of the user; and determining the users with the purchase tendency indexes meeting the preset fourth judgment conditions as target users, and providing purchase guide objects of the target commodities for the target users.
In a specific application, the purchase correlation behavior may include: search for the item and not purchase the item after a first length of time, access an item page and not purchase the item after a second length of time, place the item in a shopping cart and not purchase the item after a third length of time or enter an item checkout page and not purchase the item after a fourth length of time. The first discrimination condition may be: the weight value of the primary selection dimension is greater than a preset second threshold value or the weight value of the primary selection dimension is one of m maximum weight values; wherein m is a positive integer. The second determination condition may be: the purchasing tendency score of the user is between a preset third threshold value and a preset fourth threshold value; the purchase intention classification model is a logistic regression model, and the purchase guide object is advertisement information or a trial product.
According to the technical scheme of the embodiment of the invention, the high-weight dimension and the key dimension can be extracted according to the purchase association behavior of the user, so that the accurate positioning of the target user is realized, the user positioning strategy can be adjusted in a closed loop according to the feedback data, and the effective promotion of the commodity sales volume is finally realized.
Fig. 4 illustrates an exemplary system architecture 400 of a method of determining a target user or a device for determining a target user to which embodiments of the present invention may be applied.
As shown in fig. 4, the system architecture 400 may include terminal devices 401, 402, 403, a network 404, and a server 405 (this architecture is merely an example, and the components included in a particular architecture may be adapted according to application specific circumstances). The network 404 serves as a medium for providing communication links between the terminal devices 401, 402, 403 and the server 405. Network 404 may include various types of connections, such as wire, wireless communication links, or fiber optic cables, to name a few.
a user may use terminal devices 401, 402, 403 to interact with a server 405 over a network 404 to receive or send messages or the like. The terminal devices 401, 402, 403 may have installed thereon various communication client applications, such as shopping-like applications, web browser applications, search-like applications, instant messaging tools, mailbox clients, social platform software, etc. (by way of example only).
the terminal devices 401, 402, 403 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop portable computers, desktop computers, and the like.
The server 405 may be a server providing various services, such as a user behavior analysis server (for example only) that provides support for shopping websites browsed by users using the terminal devices 401, 402, 403. The user behavior analysis server may analyze the received user behavior data to obtain data such as high-weight dimensionality, purchase tendency score, target user, and the like (for example only).
It should be noted that the method for determining the target user provided by the embodiment of the present invention is generally executed by the server 405, and accordingly, the device for determining the target user is generally disposed in the server 405.
It should be understood that the number of terminal devices, networks, and servers in fig. 4 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.
the invention also provides the electronic equipment. The electronic device of the embodiment of the invention comprises: one or more processors; a storage device for storing one or more programs which, when executed by the one or more processors, cause the one or more processors to implement the method of determining a target user provided by the present invention.
Referring now to FIG. 5, shown is a block diagram of a computer system 500 suitable for use in implementing an electronic device of an embodiment of the present invention. The electronic device shown in fig. 5 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present invention.
As shown in fig. 5, the computer system 500 includes a Central Processing Unit (CPU)501 that can perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM)502 or a program loaded from a storage section 508 into a Random Access Memory (RAM) 503. In the RAM503, various programs and data necessary for the operation of the computer system 500 are also stored. The CPU501, ROM 502, and RAM503 are connected to each other via a bus 504. An input/output (I/O) interface 505 is also connected to bus 504.
the following components are connected to the I/O interface 505: an input portion 506 including a keyboard, a mouse, and the like; an output portion 507 including a display such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker; a storage portion 508 including a hard disk and the like; and a communication section 509 including a network interface card such as a LAN card, a modem, or the like. The communication section 509 performs communication processing via a network such as the internet. The driver 510 is also connected to the I/O interface 505 as necessary. A removable medium 511 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 510 as necessary, so that a computer program read out therefrom is mounted into the storage section 508 as necessary.
in particular, the processes described in the main step diagrams above may be implemented as computer software programs, according to embodiments of the present disclosure. For example, embodiments of the invention include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method illustrated in the main step diagram. In the above-described embodiment, the computer program can be downloaded and installed from a network through the communication section 509, and/or installed from the removable medium 511. The computer program performs the above-described functions defined in the system of the present invention when executed by the central processing unit 501.
It should be noted that the computer readable medium shown in the present invention can be a computer readable signal medium or a computer readable storage medium or any combination of the two. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, 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 disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present invention, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In the present invention, a computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, fiber optic cable, RF, etc., or any suitable combination of the foregoing.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The units described in the embodiments of the present invention may be implemented by software or hardware. The described units may also be provided in a processor, and may be described as: a processor includes a dimension weight acquisition unit and a target user location unit. Where the names of these units do not in some cases constitute a limitation on the unit itself, for example, the dimension weight obtaining unit may also be described as a "unit sending an initially selected dimension weight value to the target user location unit".
As another aspect, the present invention also provides a computer-readable medium that may be contained in the apparatus described in the above embodiments; or may be separate and not incorporated into the device. The computer readable medium carries one or more programs which, when executed by the apparatus, cause the apparatus to perform steps comprising: training a pre-established purchase intention classification model by utilizing the values of a plurality of specific users in at least one preset primary selection dimension to obtain the weight value of each primary selection dimension; wherein the specific user is a user who has performed a purchase correlation action for the target commodity and thereafter purchases the target commodity; determining a primary selection dimension with a weight value meeting a preset first judgment condition as a high-weight dimension, and acquiring a purchasing tendency score of a user who does not purchase a target commodity according to the high-weight dimension and the weight value thereof; and determining the users with the purchase tendency scores meeting the preset second judgment conditions as target users, and providing purchase guide objects of the target commodities for the target users.
According to the technical scheme of the embodiment of the invention, the high-weight dimension and the key dimension can be extracted according to the purchase association behavior of the user, so that the accurate positioning of the target user is realized, the user positioning strategy can be adjusted in a closed loop according to the feedback data, and the effective promotion of the commodity sales volume is finally realized.
The above-described embodiments should not be construed as limiting the scope of the invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions can occur, depending on design requirements and other factors. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (14)
1. A method for identifying a target user, comprising:
training a pre-established purchase intention classification model by utilizing the values of a plurality of specific users in at least one preset primary selection dimension to obtain the weight value of each primary selection dimension; wherein the specific user is a user who has performed a purchase correlation action for the target commodity and thereafter purchases the target commodity;
Determining a primary selection dimension with a weight value meeting a preset first judgment condition as a high-weight dimension, and acquiring a purchasing tendency score of a user who does not purchase a target commodity according to the high-weight dimension and the weight value thereof; and determining the users with the purchase tendency scores meeting the preset second judgment conditions as target users, and providing purchase guide objects of the target commodities for the target users.
2. The method of claim 1, wherein the primary dimension is a virtual variable; and, the method further comprises:
After determining the high weight dimension, for any user who has not purchased the destination item: and determining that the user has a high-weight dimension with a non-zero value, and taking a dimension option corresponding to the non-zero value as an portrait label of the user.
3. the method according to claim 2, wherein the obtaining of the purchase tendency score of the user who has not purchased the target product according to the high weight dimension and the weight value thereof specifically comprises:
And determining the sum of the weight values of the high-weight dimensions with non-zero values of the user as the purchase tendency score of the user.
4. The method according to claim 1, wherein the purchase guide object to which the target commodity is provided specifically includes:
Providing a target user with a purchase tendency score larger than a preset first threshold value with a purchase guide object of a target commodity;
And providing the target users with the purchase tendency scores not larger than the first threshold value with the purchase guide objects and the promotion strategies of the target commodities.
5. the method of claim 2, further comprising:
after a purchase guide object of a target commodity is provided for a determined target user, training a current purchase intention classification model by using feedback data of the target user to adjust a weight value of the primary selection dimension; wherein the feedback data comprises: receiving the value of the target user who purchases the target commodity after purchasing the guide object in the primary selection dimension;
Determining the primary dimension of which the adjusted weight value meets a preset third judgment condition as a key dimension; for any user who does not purchase a target commodity, determining the sum of the weight values of the key dimensions with non-zero values of the user as the purchase tendency index of the user; and determining the users with the purchase tendency indexes meeting the preset fourth judgment conditions as target users, and providing purchase guide objects of the target commodities for the target users.
6. The method according to any one of claims 1 to 5,
The purchase correlation behavior comprises: searching for a commodity and not purchasing the commodity after a first period of time, accessing a commodity page and not purchasing the commodity after a second period of time, placing the commodity in a shopping cart and not purchasing the commodity after a third period of time or entering a commodity settlement page and not purchasing the commodity after a fourth period of time;
The first discrimination condition includes: the weight value of the primary selection dimension is greater than a preset second threshold value or the weight value of the primary selection dimension is one of m maximum weight values; wherein m is a positive integer;
The second determination condition includes: the purchasing tendency score of the user is between a preset third threshold value and a preset fourth threshold value; and the number of the first and second groups,
The purchase intention classification model is a logistic regression model, and the purchase guide object is advertisement information or a trial product.
7. An apparatus for determining a target user, comprising:
The dimension weight obtaining unit is used for training a pre-established purchase intention classification model by utilizing the values of a plurality of specific users in at least one preset primary selection dimension to obtain the weight value of each primary selection dimension; wherein the specific user is a user who has performed a purchase correlation action for the target commodity and thereafter purchases the target commodity;
The target user positioning unit is used for determining the primary selection dimension with the weight value meeting a preset first judgment condition as a high-weight dimension, and acquiring the purchasing tendency score of any user who does not purchase a target commodity according to the high-weight dimension and the weight value of the high-weight dimension; and determining the users with the purchase tendency scores meeting the preset second judgment conditions as target users, and providing purchase guide objects of the target commodities for the target users.
8. The apparatus of claim 7, wherein the primary dimension is a virtual variable; and, the apparatus further comprises:
A portrait label determination unit for, after determining the high weight dimension, for a user of any non-purchased destination item: and determining that the user has a high-weight dimension with a non-zero value, and taking a dimension option corresponding to the non-zero value as an portrait label of the user.
9. The apparatus of claim 8, wherein the target user location unit is further configured to:
And determining the sum of the weight values of the high-weight dimensions with non-zero values of the user as the purchase tendency score of the user.
10. The apparatus of claim 7, wherein the target user location unit is further configured to:
Providing a target user with a purchase tendency score larger than a preset first threshold value with a purchase guide object of a target commodity; and providing the target users with the purchase tendency scores not larger than the first threshold value with the purchase guide objects and the promotion strategies of the target commodities.
11. The apparatus of claim 8, further comprising:
The secondary iteration unit is used for training a current purchase intention classification model by using feedback data of the target user to adjust the weight value of the primary selection dimension after a purchase guide object of a target commodity is provided for the determined target user; wherein the feedback data comprises: receiving the value of the target user who purchases the target commodity after purchasing the guide object in the primary selection dimension; determining the primary dimension of which the adjusted weight value meets a preset third judgment condition as a key dimension; for any user who does not purchase a target commodity, determining the sum of the weight values of the key dimensions with non-zero values of the user as the purchase tendency index of the user; and determining the users with the purchase tendency indexes meeting the preset fourth judgment conditions as target users, and providing purchase guide objects of the target commodities for the target users.
12. The apparatus according to any one of claims 7 to 11,
The purchase correlation behavior comprises: searching for a commodity and not purchasing the commodity after a first period of time, accessing a commodity page and not purchasing the commodity after a second period of time, placing the commodity in a shopping cart and not purchasing the commodity after a third period of time or entering a commodity settlement page and not purchasing the commodity after a fourth period of time;
The first discrimination condition includes: the weight value of the primary selection dimension is greater than a preset second threshold value or the weight value of the primary selection dimension is one of m maximum weight values; wherein m is a positive integer;
The second determination condition includes: the purchasing tendency score of the user is between a preset third threshold value and a preset fourth threshold value; and the number of the first and second groups,
The purchase intention classification model is a logistic regression model, and the purchase guide object is advertisement information or a trial product.
13. An electronic device, comprising:
One or more processors;
A storage device for storing one or more programs,
When executed by the one or more processors, cause the one or more processors to implement the method of any one of claims 1-6.
14. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the method according to any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810558184.1A CN110555747A (en) | 2018-06-01 | 2018-06-01 | method and device for determining target user |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810558184.1A CN110555747A (en) | 2018-06-01 | 2018-06-01 | method and device for determining target user |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110555747A true CN110555747A (en) | 2019-12-10 |
Family
ID=68735321
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810558184.1A Pending CN110555747A (en) | 2018-06-01 | 2018-06-01 | method and device for determining target user |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110555747A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111737587A (en) * | 2020-08-21 | 2020-10-02 | 北京每日优鲜电子商务有限公司 | Device operation method, device, electronic device and computer readable medium |
| CN113076346A (en) * | 2021-03-31 | 2021-07-06 | 北京深演智能科技股份有限公司 | Method and device for determining target object |
| CN113643068A (en) * | 2021-08-18 | 2021-11-12 | 北京小米移动软件有限公司 | Commodity purchase intention determining method and device, storage medium and electronic equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105160008A (en) * | 2015-09-21 | 2015-12-16 | 合一网络技术(北京)有限公司 | Method and device for locating suggested users |
| CN106649774A (en) * | 2016-12-27 | 2017-05-10 | 北京百度网讯科技有限公司 | Artificial intelligence-based object pushing method and apparatus |
| CN107016569A (en) * | 2017-03-21 | 2017-08-04 | 聚好看科技股份有限公司 | The targeted customer's account acquisition methods and device of a kind of networking products |
| CN107944913A (en) * | 2017-11-21 | 2018-04-20 | 重庆邮电大学 | High potential user's purchase intention Forecasting Methodology based on big data user behavior analysis |
-
2018
- 2018-06-01 CN CN201810558184.1A patent/CN110555747A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105160008A (en) * | 2015-09-21 | 2015-12-16 | 合一网络技术(北京)有限公司 | Method and device for locating suggested users |
| CN106649774A (en) * | 2016-12-27 | 2017-05-10 | 北京百度网讯科技有限公司 | Artificial intelligence-based object pushing method and apparatus |
| CN107016569A (en) * | 2017-03-21 | 2017-08-04 | 聚好看科技股份有限公司 | The targeted customer's account acquisition methods and device of a kind of networking products |
| CN107944913A (en) * | 2017-11-21 | 2018-04-20 | 重庆邮电大学 | High potential user's purchase intention Forecasting Methodology based on big data user behavior analysis |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111737587A (en) * | 2020-08-21 | 2020-10-02 | 北京每日优鲜电子商务有限公司 | Device operation method, device, electronic device and computer readable medium |
| CN113076346A (en) * | 2021-03-31 | 2021-07-06 | 北京深演智能科技股份有限公司 | Method and device for determining target object |
| CN113643068A (en) * | 2021-08-18 | 2021-11-12 | 北京小米移动软件有限公司 | Commodity purchase intention determining method and device, storage medium and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111125574B (en) | Method and device for generating information | |
| CN106897905B (en) | Method and device for pushing information and electronic equipment | |
| JP2021089742A (en) | Method for selecting advertisement, method for classifying user, and program | |
| CN110929136A (en) | Personalized recommendation method and device | |
| CN110555747A (en) | method and device for determining target user | |
| CN111782937A (en) | Information sorting method and device, electronic equipment and computer readable medium | |
| CN111695041B (en) | Method and device for recommending information | |
| WO2022156589A1 (en) | Method and device for determining live broadcast click rate | |
| CN110866625A (en) | Promotion index information generation method and device | |
| CN110992127A (en) | Article recommendation method and device | |
| CN113392920B (en) | Method, apparatus, device, medium, and program product for generating cheating prediction model | |
| CN110197317B (en) | Target user determination method and device, electronic equipment and storage medium | |
| WO2020211616A1 (en) | Method and device for processing user interaction information | |
| CN107357847B (en) | Data processing method and device | |
| CN110826823A (en) | Pricing strategy evaluation method and system | |
| CN113792952A (en) | Method and apparatus for generating a model | |
| CN113592315A (en) | Method and device for processing dispute order | |
| CN110110199B (en) | Information output method and device | |
| CN113742564A (en) | Target resource pushing method and device | |
| CN112784861A (en) | Similarity determination method and device, electronic equipment and storage medium | |
| CN109993559A (en) | Model training method and system | |
| CN113066479A (en) | Method and device for evaluating model | |
| CN108846743B (en) | Method and apparatus for generating information | |
| CN110517055B (en) | Method and device for generating physical place promotion information | |
| CN108628861B (en) | Method and device for pushing information |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |